From f239aed5e674965691846e8ce3f187dd47523689 Mon Sep 17 00:00:00 2001 From: Luca Boccassi Date: Wed, 16 Aug 2017 18:42:05 +0100 Subject: New upstream version 17.08 Change-Id: I288b50990f52646089d6b1f3aaa6ba2f091a51d7 Signed-off-by: Luca Boccassi --- lib/librte_ether/Makefile | 5 +- lib/librte_ether/rte_ethdev.c | 261 ++--- lib/librte_ether/rte_ethdev.h | 347 ++---- lib/librte_ether/rte_ethdev_pci.h | 11 +- lib/librte_ether/rte_ethdev_vdev.h | 1 - lib/librte_ether/rte_ether_version.map | 51 +- lib/librte_ether/rte_flow.c | 249 +++++ lib/librte_ether/rte_flow.h | 176 ++- lib/librte_ether/rte_flow_driver.h | 5 + lib/librte_ether/rte_tm.c | 438 ++++++++ lib/librte_ether/rte_tm.h | 1912 ++++++++++++++++++++++++++++++++ lib/librte_ether/rte_tm_driver.h | 366 ++++++ 12 files changed, 3355 insertions(+), 467 deletions(-) create mode 100644 lib/librte_ether/rte_tm.c create mode 100644 lib/librte_ether/rte_tm.h create mode 100644 lib/librte_ether/rte_tm_driver.h (limited to 'lib/librte_ether') diff --git a/lib/librte_ether/Makefile b/lib/librte_ether/Makefile index 93fdde10..db692ae4 100644 --- a/lib/librte_ether/Makefile +++ b/lib/librte_ether/Makefile @@ -1,6 +1,6 @@ # BSD LICENSE # -# Copyright(c) 2010-2016 Intel Corporation. All rights reserved. +# Copyright(c) 2010-2017 Intel Corporation. All rights reserved. # All rights reserved. # # Redistribution and use in source and binary forms, with or without @@ -45,6 +45,7 @@ LIBABIVER := 6 SRCS-y += rte_ethdev.c SRCS-y += rte_flow.c +SRCS-y += rte_tm.c # # Export include files @@ -56,5 +57,7 @@ SYMLINK-y-include += rte_eth_ctrl.h SYMLINK-y-include += rte_dev_info.h SYMLINK-y-include += rte_flow.h SYMLINK-y-include += rte_flow_driver.h +SYMLINK-y-include += rte_tm.h +SYMLINK-y-include += rte_tm_driver.h include $(RTE_SDK)/mk/rte.lib.mk diff --git a/lib/librte_ether/rte_ethdev.c b/lib/librte_ether/rte_ethdev.c index 83898a8f..0597641e 100644 --- a/lib/librte_ether/rte_ethdev.c +++ b/lib/librte_ether/rte_ethdev.c @@ -1,7 +1,7 @@ /*- * BSD LICENSE * - * Copyright(c) 2010-2016 Intel Corporation. All rights reserved. + * Copyright(c) 2010-2017 Intel Corporation. All rights reserved. * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -72,7 +72,6 @@ static const char *MZ_RTE_ETH_DEV_DATA = "rte_eth_dev_data"; struct rte_eth_dev rte_eth_devices[RTE_MAX_ETHPORTS]; static struct rte_eth_dev_data *rte_eth_dev_data; static uint8_t eth_dev_last_created_port; -static uint8_t nb_ports; /* spinlock for eth device callbacks */ static rte_spinlock_t rte_eth_dev_cb_lock = RTE_SPINLOCK_INITIALIZER; @@ -129,6 +128,7 @@ struct rte_eth_dev_callback { TAILQ_ENTRY(rte_eth_dev_callback) next; /**< Callbacks list */ rte_eth_dev_cb_fn cb_fn; /**< Callback address */ void *cb_arg; /**< Parameter for callback */ + void *ret_param; /**< Return parameter */ enum rte_eth_event_type event; /**< Interrupt event type */ uint32_t active; /**< Callback is executing */ }; @@ -178,9 +178,11 @@ rte_eth_dev_allocated(const char *name) unsigned i; for (i = 0; i < RTE_MAX_ETHPORTS; i++) { - if ((rte_eth_devices[i].state == RTE_ETH_DEV_ATTACHED) && - strcmp(rte_eth_devices[i].data->name, name) == 0) - return &rte_eth_devices[i]; + if (rte_eth_devices[i].state == RTE_ETH_DEV_ATTACHED && + rte_eth_devices[i].device) { + if (!strcmp(rte_eth_devices[i].device->name, name)) + return &rte_eth_devices[i]; + } } return NULL; } @@ -207,7 +209,6 @@ eth_dev_get(uint8_t port_id) TAILQ_INIT(&(eth_dev->link_intr_cbs)); eth_dev_last_created_port = port_id; - nb_ports++; return eth_dev; } @@ -280,7 +281,6 @@ rte_eth_dev_release_port(struct rte_eth_dev *eth_dev) return -EINVAL; eth_dev->state = RTE_ETH_DEV_UNUSED; - nb_ports--; return 0; } @@ -288,7 +288,8 @@ int rte_eth_dev_is_valid_port(uint8_t port_id) { if (port_id >= RTE_MAX_ETHPORTS || - rte_eth_devices[port_id].state != RTE_ETH_DEV_ATTACHED) + (rte_eth_devices[port_id].state != RTE_ETH_DEV_ATTACHED && + rte_eth_devices[port_id].state != RTE_ETH_DEV_DEFERRED)) return 0; else return 1; @@ -304,13 +305,21 @@ rte_eth_dev_socket_id(uint8_t port_id) uint8_t rte_eth_dev_count(void) { - return nb_ports; + uint8_t p; + uint8_t count; + + count = 0; + + RTE_ETH_FOREACH_DEV(p) + count++; + + return count; } int rte_eth_dev_get_name_by_port(uint8_t port_id, char *name) { - char *tmp; + const char *tmp; RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -EINVAL); @@ -321,7 +330,7 @@ rte_eth_dev_get_name_by_port(uint8_t port_id, char *name) /* shouldn't check 'rte_eth_devices[i].data', * because it might be overwritten by VDEV PMD */ - tmp = rte_eth_dev_data[port_id].name; + tmp = rte_eth_devices[port_id].device->name; strcpy(name, tmp); return 0; } @@ -329,6 +338,7 @@ rte_eth_dev_get_name_by_port(uint8_t port_id, char *name) int rte_eth_dev_get_port_by_name(const char *name, uint8_t *port_id) { + int ret; int i; if (name == NULL) { @@ -336,16 +346,14 @@ rte_eth_dev_get_port_by_name(const char *name, uint8_t *port_id) return -EINVAL; } - if (!nb_ports) - return -ENODEV; - - *port_id = RTE_MAX_ETHPORTS; RTE_ETH_FOREACH_DEV(i) { - if (!strncmp(name, - rte_eth_dev_data[i].name, strlen(name))) { + if (!rte_eth_devices[i].device) + continue; + ret = strncmp(name, rte_eth_devices[i].device->name, + strlen(name)); + if (ret == 0) { *port_id = i; - return 0; } } @@ -359,16 +367,6 @@ rte_eth_dev_is_detachable(uint8_t port_id) RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -EINVAL); - switch (rte_eth_devices[port_id].data->kdrv) { - case RTE_KDRV_IGB_UIO: - case RTE_KDRV_UIO_GENERIC: - case RTE_KDRV_NIC_UIO: - case RTE_KDRV_NONE: - case RTE_KDRV_VFIO: - break; - default: - return -ENOTSUP; - } dev_flags = rte_eth_devices[port_id].data->dev_flags; if ((dev_flags & RTE_ETH_DEV_DETACHABLE) && (!(dev_flags & RTE_ETH_DEV_BONDED_SLAVE))) @@ -438,12 +436,14 @@ rte_eth_dev_detach(uint8_t port_id, char *name) if (rte_eth_dev_is_detachable(port_id)) goto err; - snprintf(name, sizeof(rte_eth_devices[port_id].data->name), - "%s", rte_eth_devices[port_id].data->name); - ret = rte_eal_dev_detach(name); + snprintf(name, RTE_DEV_NAME_MAX_LEN, "%s", + rte_eth_devices[port_id].device->name); + + ret = rte_eal_dev_detach(rte_eth_devices[port_id].device); if (ret < 0) goto err; + rte_eth_devices[port_id].state = RTE_ETH_DEV_UNUSED; return 0; err: @@ -753,13 +753,13 @@ rte_eth_dev_configure(uint8_t port_id, uint16_t nb_rx_q, uint16_t nb_tx_q, if ((dev_conf->intr_conf.lsc == 1) && (!(dev->data->dev_flags & RTE_ETH_DEV_INTR_LSC))) { RTE_PMD_DEBUG_TRACE("driver %s does not support lsc\n", - dev->data->drv_name); + dev->device->driver->name); return -EINVAL; } if ((dev_conf->intr_conf.rmv == 1) && (!(dev->data->dev_flags & RTE_ETH_DEV_INTR_RMV))) { RTE_PMD_DEBUG_TRACE("driver %s does not support rmv\n", - dev->data->drv_name); + dev->device->driver->name); return -EINVAL; } @@ -1900,7 +1900,7 @@ rte_eth_dev_info_get(uint8_t port_id, struct rte_eth_dev_info *dev_info) RTE_FUNC_PTR_OR_RET(*dev->dev_ops->dev_infos_get); (*dev->dev_ops->dev_infos_get)(dev, dev_info); - dev_info->driver_name = dev->data->drv_name; + dev_info->driver_name = dev->device->driver->name; dev_info->nb_rx_queues = dev->data->nb_rx_queues; dev_info->nb_tx_queues = dev->data->nb_tx_queues; } @@ -1975,6 +1975,7 @@ int rte_eth_dev_vlan_filter(uint8_t port_id, uint16_t vlan_id, int on) { struct rte_eth_dev *dev; + int ret; RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); dev = &rte_eth_devices[port_id]; @@ -1990,7 +1991,23 @@ rte_eth_dev_vlan_filter(uint8_t port_id, uint16_t vlan_id, int on) } RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->vlan_filter_set, -ENOTSUP); - return (*dev->dev_ops->vlan_filter_set)(dev, vlan_id, on); + ret = (*dev->dev_ops->vlan_filter_set)(dev, vlan_id, on); + if (ret == 0) { + struct rte_vlan_filter_conf *vfc; + int vidx; + int vbit; + + vfc = &dev->data->vlan_filter_conf; + vidx = vlan_id / 64; + vbit = vlan_id % 64; + + if (on) + vfc->ids[vidx] |= UINT64_C(1) << vbit; + else + vfc->ids[vidx] &= ~(UINT64_C(1) << vbit); + } + + return ret; } int @@ -2351,6 +2368,7 @@ get_mac_addr_index(uint8_t port_id, const struct ether_addr *addr) struct rte_eth_dev *dev = &rte_eth_devices[port_id]; unsigned i; + RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); rte_eth_dev_info_get(port_id, &dev_info); for (i = 0; i < dev_info.max_mac_addrs; i++) @@ -2718,12 +2736,13 @@ rte_eth_dev_callback_unregister(uint8_t port_id, return ret; } -void +int _rte_eth_dev_callback_process(struct rte_eth_dev *dev, - enum rte_eth_event_type event, void *cb_arg) + enum rte_eth_event_type event, void *cb_arg, void *ret_param) { struct rte_eth_dev_callback *cb_lst; struct rte_eth_dev_callback dev_cb; + int rc = 0; rte_spinlock_lock(&rte_eth_dev_cb_lock); TAILQ_FOREACH(cb_lst, &(dev->link_intr_cbs), next) { @@ -2733,14 +2752,17 @@ _rte_eth_dev_callback_process(struct rte_eth_dev *dev, cb_lst->active = 1; if (cb_arg != NULL) dev_cb.cb_arg = cb_arg; + if (ret_param != NULL) + dev_cb.ret_param = ret_param; rte_spinlock_unlock(&rte_eth_dev_cb_lock); - dev_cb.cb_fn(dev->data->port_id, dev_cb.event, - dev_cb.cb_arg); + rc = dev_cb.cb_fn(dev->data->port_id, dev_cb.event, + dev_cb.cb_arg, dev_cb.ret_param); rte_spinlock_lock(&rte_eth_dev_cb_lock); cb_lst->active = 0; } rte_spinlock_unlock(&rte_eth_dev_cb_lock); + return rc; } int @@ -2789,7 +2811,7 @@ rte_eth_dma_zone_reserve(const struct rte_eth_dev *dev, const char *ring_name, const struct rte_memzone *mz; snprintf(z_name, sizeof(z_name), "%s_%s_%d_%d", - dev->data->drv_name, ring_name, + dev->device->driver->name, ring_name, dev->data->port_id, queue_id); mz = rte_memzone_lookup(z_name); @@ -2872,128 +2894,6 @@ rte_eth_dev_rx_intr_disable(uint8_t port_id, return (*dev->dev_ops->rx_queue_intr_disable)(dev, queue_id); } -#ifdef RTE_NIC_BYPASS -int rte_eth_dev_bypass_init(uint8_t port_id) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_init, -ENOTSUP); - (*dev->dev_ops->bypass_init)(dev); - return 0; -} - -int -rte_eth_dev_bypass_state_show(uint8_t port_id, uint32_t *state) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_state_show, -ENOTSUP); - (*dev->dev_ops->bypass_state_show)(dev, state); - return 0; -} - -int -rte_eth_dev_bypass_state_set(uint8_t port_id, uint32_t *new_state) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_state_set, -ENOTSUP); - (*dev->dev_ops->bypass_state_set)(dev, new_state); - return 0; -} - -int -rte_eth_dev_bypass_event_show(uint8_t port_id, uint32_t event, uint32_t *state) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_state_show, -ENOTSUP); - (*dev->dev_ops->bypass_event_show)(dev, event, state); - return 0; -} - -int -rte_eth_dev_bypass_event_store(uint8_t port_id, uint32_t event, uint32_t state) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_event_set, -ENOTSUP); - (*dev->dev_ops->bypass_event_set)(dev, event, state); - return 0; -} - -int -rte_eth_dev_wd_timeout_store(uint8_t port_id, uint32_t timeout) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_wd_timeout_set, -ENOTSUP); - (*dev->dev_ops->bypass_wd_timeout_set)(dev, timeout); - return 0; -} - -int -rte_eth_dev_bypass_ver_show(uint8_t port_id, uint32_t *ver) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_ver_show, -ENOTSUP); - (*dev->dev_ops->bypass_ver_show)(dev, ver); - return 0; -} - -int -rte_eth_dev_bypass_wd_timeout_show(uint8_t port_id, uint32_t *wd_timeout) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_wd_timeout_show, -ENOTSUP); - (*dev->dev_ops->bypass_wd_timeout_show)(dev, wd_timeout); - return 0; -} - -int -rte_eth_dev_bypass_wd_reset(uint8_t port_id) -{ - struct rte_eth_dev *dev; - - RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); - - dev = &rte_eth_devices[port_id]; - - RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->bypass_wd_reset, -ENOTSUP); - (*dev->dev_ops->bypass_wd_reset)(dev); - return 0; -} -#endif int rte_eth_dev_filter_supported(uint8_t port_id, enum rte_filter_type filter_type) @@ -3472,3 +3372,40 @@ rte_eth_dev_l2_tunnel_offload_set(uint8_t port_id, -ENOTSUP); return (*dev->dev_ops->l2_tunnel_offload_set)(dev, l2_tunnel, mask, en); } + +static void +rte_eth_dev_adjust_nb_desc(uint16_t *nb_desc, + const struct rte_eth_desc_lim *desc_lim) +{ + if (desc_lim->nb_align != 0) + *nb_desc = RTE_ALIGN_CEIL(*nb_desc, desc_lim->nb_align); + + if (desc_lim->nb_max != 0) + *nb_desc = RTE_MIN(*nb_desc, desc_lim->nb_max); + + *nb_desc = RTE_MAX(*nb_desc, desc_lim->nb_min); +} + +int +rte_eth_dev_adjust_nb_rx_tx_desc(uint8_t port_id, + uint16_t *nb_rx_desc, + uint16_t *nb_tx_desc) +{ + struct rte_eth_dev *dev; + struct rte_eth_dev_info dev_info; + + RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV); + + dev = &rte_eth_devices[port_id]; + RTE_FUNC_PTR_OR_ERR_RET(*dev->dev_ops->dev_infos_get, -ENOTSUP); + + rte_eth_dev_info_get(port_id, &dev_info); + + if (nb_rx_desc != NULL) + rte_eth_dev_adjust_nb_desc(nb_rx_desc, &dev_info.rx_desc_lim); + + if (nb_tx_desc != NULL) + rte_eth_dev_adjust_nb_desc(nb_tx_desc, &dev_info.tx_desc_lim); + + return 0; +} diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h index 0f38b45f..0adf3274 100644 --- a/lib/librte_ether/rte_ethdev.h +++ b/lib/librte_ether/rte_ethdev.h @@ -1,7 +1,7 @@ /*- * BSD LICENSE * - * Copyright(c) 2010-2016 Intel Corporation. All rights reserved. + * Copyright(c) 2010-2017 Intel Corporation. All rights reserved. * All rights reserved. * * Redistribution and use in source and binary forms, with or without @@ -118,7 +118,7 @@ * - NIC queue statistics mappings * * Any other configuration will not be stored and will need to be re-entered - * after a call to rte_eth_dev_start(). + * before a call to rte_eth_dev_start(). * * Finally, a network application can close an Ethernet device by invoking the * rte_eth_dev_close() function. @@ -172,8 +172,6 @@ extern "C" { #include -#include - /* Use this macro to check if LRO API is supported */ #define RTE_ETHDEV_HAS_LRO_SUPPORT @@ -373,6 +371,14 @@ enum rte_vlan_type { ETH_VLAN_TYPE_MAX, }; +/** + * A structure used to describe a vlan filter. + * If the bit corresponding to a VID is set, such VID is on. + */ +struct rte_vlan_filter_conf { + uint64_t ids[64]; +}; + /** * A structure used to configure the Receive Side Scaling (RSS) feature * of an Ethernet port. @@ -629,6 +635,24 @@ struct rte_eth_vmdq_dcb_conf { /**< Selects a queue in a pool */ }; +/** + * A structure used to configure the VMDQ feature of an Ethernet port when + * not combined with the DCB feature. + * + * Using this feature, packets are routed to a pool of queues. By default, + * the pool selection is based on the MAC address, the vlan id in the + * vlan tag as specified in the pool_map array. + * Passing the ETH_VMDQ_ACCEPT_UNTAG in the rx_mode field allows pool + * selection using only the MAC address. MAC address to pool mapping is done + * using the rte_eth_dev_mac_addr_add function, with the pool parameter + * corresponding to the pool id. + * + * Queue selection within the selected pool will be done using RSS when + * it is enabled or revert to the first queue of the pool if not. + * + * A default pool may be used, if desired, to route all traffic which + * does not match the vlan filter rules or any pool MAC address. + */ struct rte_eth_vmdq_rx_conf { enum rte_eth_nb_pools nb_queue_pools; /**< VMDq only mode, 8 or 64 pools */ uint8_t enable_default_pool; /**< If non-zero, use a default pool */ @@ -901,6 +925,10 @@ struct rte_eth_conf { #define DEV_TX_OFFLOAD_IPIP_TNL_TSO 0x00000800 /**< Used for tunneling packet. */ #define DEV_TX_OFFLOAD_GENEVE_TNL_TSO 0x00001000 /**< Used for tunneling packet. */ #define DEV_TX_OFFLOAD_MACSEC_INSERT 0x00002000 +#define DEV_TX_OFFLOAD_MT_LOCKFREE 0x00004000 +/**< Multiple threads can invoke rte_eth_tx_burst() concurrently on the same + * tx queue without SW lock. + */ struct rte_pci_device; @@ -1048,6 +1076,8 @@ TAILQ_HEAD(rte_eth_dev_cb_list, rte_eth_dev_callback); } \ } while (0) +#define RTE_ETH_DEV_TO_PCI(eth_dev) RTE_DEV_TO_PCI((eth_dev)->device) + /** * l2 tunnel configuration. */ @@ -1381,59 +1411,6 @@ typedef int (*eth_l2_tunnel_offload_set_t) uint8_t en); /**< @internal enable/disable the l2 tunnel offload functions */ -#ifdef RTE_NIC_BYPASS - -enum { - RTE_BYPASS_MODE_NONE, - RTE_BYPASS_MODE_NORMAL, - RTE_BYPASS_MODE_BYPASS, - RTE_BYPASS_MODE_ISOLATE, - RTE_BYPASS_MODE_NUM, -}; - -#define RTE_BYPASS_MODE_VALID(x) \ - ((x) > RTE_BYPASS_MODE_NONE && (x) < RTE_BYPASS_MODE_NUM) - -enum { - RTE_BYPASS_EVENT_NONE, - RTE_BYPASS_EVENT_START, - RTE_BYPASS_EVENT_OS_ON = RTE_BYPASS_EVENT_START, - RTE_BYPASS_EVENT_POWER_ON, - RTE_BYPASS_EVENT_OS_OFF, - RTE_BYPASS_EVENT_POWER_OFF, - RTE_BYPASS_EVENT_TIMEOUT, - RTE_BYPASS_EVENT_NUM -}; - -#define RTE_BYPASS_EVENT_VALID(x) \ - ((x) > RTE_BYPASS_EVENT_NONE && (x) < RTE_BYPASS_MODE_NUM) - -enum { - RTE_BYPASS_TMT_OFF, /* timeout disabled. */ - RTE_BYPASS_TMT_1_5_SEC, /* timeout for 1.5 seconds */ - RTE_BYPASS_TMT_2_SEC, /* timeout for 2 seconds */ - RTE_BYPASS_TMT_3_SEC, /* timeout for 3 seconds */ - RTE_BYPASS_TMT_4_SEC, /* timeout for 4 seconds */ - RTE_BYPASS_TMT_8_SEC, /* timeout for 8 seconds */ - RTE_BYPASS_TMT_16_SEC, /* timeout for 16 seconds */ - RTE_BYPASS_TMT_32_SEC, /* timeout for 32 seconds */ - RTE_BYPASS_TMT_NUM -}; - -#define RTE_BYPASS_TMT_VALID(x) \ - ((x) == RTE_BYPASS_TMT_OFF || \ - ((x) > RTE_BYPASS_TMT_OFF && (x) < RTE_BYPASS_TMT_NUM)) - -typedef void (*bypass_init_t)(struct rte_eth_dev *dev); -typedef int32_t (*bypass_state_set_t)(struct rte_eth_dev *dev, uint32_t *new_state); -typedef int32_t (*bypass_state_show_t)(struct rte_eth_dev *dev, uint32_t *state); -typedef int32_t (*bypass_event_set_t)(struct rte_eth_dev *dev, uint32_t state, uint32_t event); -typedef int32_t (*bypass_event_show_t)(struct rte_eth_dev *dev, uint32_t event_shift, uint32_t *event); -typedef int32_t (*bypass_wd_timeout_set_t)(struct rte_eth_dev *dev, uint32_t timeout); -typedef int32_t (*bypass_wd_timeout_show_t)(struct rte_eth_dev *dev, uint32_t *wd_timeout); -typedef int32_t (*bypass_ver_show_t)(struct rte_eth_dev *dev, uint32_t *ver); -typedef int32_t (*bypass_wd_reset_t)(struct rte_eth_dev *dev); -#endif typedef int (*eth_filter_ctrl_t)(struct rte_eth_dev *dev, enum rte_filter_type filter_type, @@ -1441,6 +1418,9 @@ typedef int (*eth_filter_ctrl_t)(struct rte_eth_dev *dev, void *arg); /**< @internal Take operations to assigned filter type on an Ethernet device */ +typedef int (*eth_tm_ops_get_t)(struct rte_eth_dev *dev, void *ops); +/**< @internal Get Traffic Management (TM) operations on an Ethernet device */ + typedef int (*eth_get_dcb_info)(struct rte_eth_dev *dev, struct rte_eth_dcb_info *dcb_info); /**< @internal Get dcb information on an Ethernet device */ @@ -1460,7 +1440,7 @@ struct eth_dev_ops { eth_promiscuous_enable_t promiscuous_enable; /**< Promiscuous ON. */ eth_promiscuous_disable_t promiscuous_disable;/**< Promiscuous OFF. */ eth_allmulticast_enable_t allmulticast_enable;/**< RX multicast ON. */ - eth_allmulticast_disable_t allmulticast_disable;/**< RX multicast OF. */ + eth_allmulticast_disable_t allmulticast_disable;/**< RX multicast OFF. */ eth_mac_addr_remove_t mac_addr_remove; /**< Remove MAC address. */ eth_mac_addr_add_t mac_addr_add; /**< Add a MAC address. */ eth_mac_addr_set_t mac_addr_set; /**< Set a MAC address. */ @@ -1540,18 +1520,6 @@ struct eth_dev_ops { eth_get_eeprom_t get_eeprom; /**< Get eeprom data. */ eth_set_eeprom_t set_eeprom; /**< Set eeprom. */ - /* bypass control */ -#ifdef RTE_NIC_BYPASS - bypass_init_t bypass_init; - bypass_state_set_t bypass_state_set; - bypass_state_show_t bypass_state_show; - bypass_event_set_t bypass_event_set; - bypass_event_show_t bypass_event_show; - bypass_wd_timeout_set_t bypass_wd_timeout_set; - bypass_wd_timeout_show_t bypass_wd_timeout_show; - bypass_ver_show_t bypass_ver_show; - bypass_wd_reset_t bypass_wd_reset; -#endif eth_filter_ctrl_t filter_ctrl; /**< common filter control. */ @@ -1573,6 +1541,9 @@ struct eth_dev_ops { /**< Get extended device statistic values by ID. */ eth_xstats_get_names_by_id_t xstats_get_names_by_id; /**< Get name of extended device statistics by ID. */ + + eth_tm_ops_get_t tm_ops_get; + /**< Get Traffic Management (TM) operations. */ }; /** @@ -1644,6 +1615,7 @@ struct rte_eth_rxtx_callback { enum rte_eth_dev_state { RTE_ETH_DEV_UNUSED = 0, RTE_ETH_DEV_ATTACHED, + RTE_ETH_DEV_DEFERRED, }; /** @@ -1687,7 +1659,7 @@ struct rte_eth_dev_sriov { }; #define RTE_ETH_DEV_SRIOV(dev) ((dev)->data->sriov) -#define RTE_ETH_NAME_MAX_LEN (32) +#define RTE_ETH_NAME_MAX_LEN RTE_DEV_NAME_MAX_LEN /** * @internal @@ -1737,7 +1709,8 @@ struct rte_eth_dev_data { uint32_t dev_flags; /**< Capabilities */ enum rte_kernel_driver kdrv; /**< Kernel driver passthrough */ int numa_node; /**< NUMA node connection */ - const char *drv_name; /**< Driver name */ + struct rte_vlan_filter_conf vlan_filter_conf; + /**< VLAN filter configuration. */ }; /** Device supports hotplug detach */ @@ -1777,13 +1750,12 @@ uint8_t rte_eth_find_next(uint8_t port_id); /** * Get the total number of Ethernet devices that have been successfully - * initialized by the [matching] Ethernet driver during the PCI probing phase. - * All devices whose port identifier is in the range - * [0, rte_eth_dev_count() - 1] can be operated on by network applications - * immediately after invoking rte_eal_init(). - * If the application unplugs a port using hotplug function, The enabled port - * numbers may be noncontiguous. In the case, the applications need to manage - * enabled port by using the ``RTE_ETH_FOREACH_DEV()`` macro. + * initialized by the matching Ethernet driver during the PCI probing phase + * and that are available for applications to use. These devices must be + * accessed by using the ``RTE_ETH_FOREACH_DEV()`` macro to deal with + * non-contiguous ranges of devices. + * These non-contiguous ranges can be created by calls to hotplug functions or + * by some PMDs. * * @return * - The total number of usable Ethernet devices. @@ -1859,7 +1831,8 @@ int rte_eth_dev_attach(const char *devargs, uint8_t *port_id); * @param port_id * The port identifier of the device to detach. * @param devname - * A pointer to a device name actually detached. + * A pointer to a buffer that will be filled with the device name. + * This buffer must be at least RTE_DEV_NAME_MAX_LEN long. * @return * 0 on success and devname is filled, negative on error */ @@ -2358,7 +2331,7 @@ rte_eth_xstats_get_names_by_id(uint8_t port_id, * @param port_id * The port identifier of the Ethernet device. * @param ids - * A pointer to an ids array passed by application. This tells wich + * A pointer to an ids array passed by application. This tells which * statistics values function should retrieve. This parameter * can be set to NULL if n is 0. In this case function will retrieve * all avalible statistics. @@ -2997,6 +2970,10 @@ static inline int rte_eth_tx_descriptor_status(uint8_t port_id, * rte_eth_tx_burst() function must [attempt to] free the *rte_mbuf* buffers * of those packets whose transmission was effectively completed. * + * If the PMD is DEV_TX_OFFLOAD_MT_LOCKFREE capable, multiple threads can + * invoke this function concurrently on the same tx queue without SW lock. + * @see rte_eth_dev_info_get, struct rte_eth_txconf::txq_flags + * * @param port_id * The port identifier of the Ethernet device. * @param queue_id @@ -3266,7 +3243,7 @@ rte_eth_tx_buffer_flush(uint8_t port_id, uint16_t queue_id, * causing N packets to be sent, and the error callback to be called for * the rest. */ -static inline uint16_t __attribute__((always_inline)) +static __rte_always_inline uint16_t rte_eth_tx_buffer(uint8_t port_id, uint16_t queue_id, struct rte_eth_dev_tx_buffer *buffer, struct rte_mbuf *tx_pkt) { @@ -3401,8 +3378,8 @@ enum rte_eth_event_type { RTE_ETH_EVENT_MAX /**< max value of this enum */ }; -typedef void (*rte_eth_dev_cb_fn)(uint8_t port_id, \ - enum rte_eth_event_type event, void *cb_arg); +typedef int (*rte_eth_dev_cb_fn)(uint8_t port_id, + enum rte_eth_event_type event, void *cb_arg, void *ret_param); /**< user application callback to be registered for interrupts */ @@ -3419,11 +3396,6 @@ typedef void (*rte_eth_dev_cb_fn)(uint8_t port_id, \ * @param cb_arg * Pointer to the parameters for the registered callback. * - * The user data is overwritten in the case of RTE_ETH_EVENT_VF_MBOX. - * This even occurs when a message from the VF is received by the PF. - * The user data is overwritten with struct rte_pmd_ixgbe_mb_event_param. - * This struct is defined in rte_pmd_ixgbe.h. - * * @return * - On success, zero. * - On failure, a negative value. @@ -3463,15 +3435,17 @@ int rte_eth_dev_callback_unregister(uint8_t port_id, * @param event * Eth device interrupt event type. * @param cb_arg - * Update callback parameter to pass data back to user application. + * callback parameter. + * @param ret_param + * To pass data back to user application. * This allows the user application to decide if a particular function * is permitted or not. * * @return - * void + * int */ -void _rte_eth_dev_callback_process(struct rte_eth_dev *dev, - enum rte_eth_event_type event, void *cb_arg); +int _rte_eth_dev_callback_process(struct rte_eth_dev *dev, + enum rte_eth_event_type event, void *cb_arg, void *ret_param); /** * When there is no rx packet coming in Rx Queue for a long time, we can @@ -3827,171 +3801,6 @@ int rte_eth_mirror_rule_reset(uint8_t port_id, int rte_eth_set_queue_rate_limit(uint8_t port_id, uint16_t queue_idx, uint16_t tx_rate); -/** - * Initialize bypass logic. This function needs to be called before - * executing any other bypass API. - * - * @param port - * The port identifier of the Ethernet device. - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_bypass_init(uint8_t port); - -/** - * Return bypass state. - * - * @param port - * The port identifier of the Ethernet device. - * @param state - * The return bypass state. - * - (1) Normal mode - * - (2) Bypass mode - * - (3) Isolate mode - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_bypass_state_show(uint8_t port, uint32_t *state); - -/** - * Set bypass state - * - * @param port - * The port identifier of the Ethernet device. - * @param new_state - * The current bypass state. - * - (1) Normal mode - * - (2) Bypass mode - * - (3) Isolate mode - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_bypass_state_set(uint8_t port, uint32_t *new_state); - -/** - * Return bypass state when given event occurs. - * - * @param port - * The port identifier of the Ethernet device. - * @param event - * The bypass event - * - (1) Main power on (power button is pushed) - * - (2) Auxiliary power on (power supply is being plugged) - * - (3) Main power off (system shutdown and power supply is left plugged in) - * - (4) Auxiliary power off (power supply is being unplugged) - * - (5) Display or set the watchdog timer - * @param state - * The bypass state when given event occurred. - * - (1) Normal mode - * - (2) Bypass mode - * - (3) Isolate mode - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_bypass_event_show(uint8_t port, uint32_t event, uint32_t *state); - -/** - * Set bypass state when given event occurs. - * - * @param port - * The port identifier of the Ethernet device. - * @param event - * The bypass event - * - (1) Main power on (power button is pushed) - * - (2) Auxiliary power on (power supply is being plugged) - * - (3) Main power off (system shutdown and power supply is left plugged in) - * - (4) Auxiliary power off (power supply is being unplugged) - * - (5) Display or set the watchdog timer - * @param state - * The assigned state when given event occurs. - * - (1) Normal mode - * - (2) Bypass mode - * - (3) Isolate mode - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_bypass_event_store(uint8_t port, uint32_t event, uint32_t state); - -/** - * Set bypass watchdog timeout count. - * - * @param port - * The port identifier of the Ethernet device. - * @param timeout - * The timeout to be set. - * - (0) 0 seconds (timer is off) - * - (1) 1.5 seconds - * - (2) 2 seconds - * - (3) 3 seconds - * - (4) 4 seconds - * - (5) 8 seconds - * - (6) 16 seconds - * - (7) 32 seconds - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_wd_timeout_store(uint8_t port, uint32_t timeout); - -/** - * Get bypass firmware version. - * - * @param port - * The port identifier of the Ethernet device. - * @param ver - * The firmware version - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_bypass_ver_show(uint8_t port, uint32_t *ver); - -/** - * Return bypass watchdog timeout in seconds - * - * @param port - * The port identifier of the Ethernet device. - * @param wd_timeout - * The return watchdog timeout. "0" represents timer expired - * - (0) 0 seconds (timer is off) - * - (1) 1.5 seconds - * - (2) 2 seconds - * - (3) 3 seconds - * - (4) 4 seconds - * - (5) 8 seconds - * - (6) 16 seconds - * - (7) 32 seconds - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_bypass_wd_timeout_show(uint8_t port, uint32_t *wd_timeout); - -/** - * Reset bypass watchdog timer - * - * @param port - * The port identifier of the Ethernet device. - * @return - * - (0) if successful. - * - (-ENOTSUP) if hardware doesn't support. - * - (-EINVAL) if bad parameter. - */ -int rte_eth_dev_bypass_wd_reset(uint8_t port); - /** * Configuration of Receive Side Scaling hash computation of Ethernet device. * @@ -4587,7 +4396,7 @@ rte_eth_dev_l2_tunnel_offload_set(uint8_t port_id, * @param port_id * pointer to port identifier of the device * @return -* - (0) if successful. +* - (0) if successful and port_id is filled. * - (-ENODEV or -EINVAL) on failure. */ int @@ -4607,6 +4416,26 @@ rte_eth_dev_get_port_by_name(const char *name, uint8_t *port_id); int rte_eth_dev_get_name_by_port(uint8_t port_id, char *name); +/** + * Check that numbers of Rx and Tx descriptors satisfy descriptors limits from + * the ethernet device information, otherwise adjust them to boundaries. + * + * @param port_id + * The port identifier of the Ethernet device. + * @param nb_rx_desc + * A pointer to a uint16_t where the number of receive + * descriptors stored. + * @param nb_tx_desc + * A pointer to a uint16_t where the number of transmit + * descriptors stored. + * @return + * - (0) if successful. + * - (-ENOTSUP, -ENODEV or -EINVAL) on failure. + */ +int rte_eth_dev_adjust_nb_rx_tx_desc(uint8_t port_id, + uint16_t *nb_rx_desc, + uint16_t *nb_tx_desc); + #ifdef __cplusplus } #endif diff --git a/lib/librte_ether/rte_ethdev_pci.h b/lib/librte_ether/rte_ethdev_pci.h index d3bc03cf..56b10721 100644 --- a/lib/librte_ether/rte_ethdev_pci.h +++ b/lib/librte_ether/rte_ethdev_pci.h @@ -45,9 +45,6 @@ * The *eth_dev* pointer is the address of the *rte_eth_dev* structure. * @param pci_dev * The *pci_dev* pointer is the address of the *rte_pci_device* structure. - * - * @return - * - 0 on success, negative on error */ static inline void rte_eth_copy_pci_info(struct rte_eth_dev *eth_dev, @@ -69,7 +66,6 @@ rte_eth_copy_pci_info(struct rte_eth_dev *eth_dev, eth_dev->data->kdrv = pci_dev->kdrv; eth_dev->data->numa_node = pci_dev->device.numa_node; - eth_dev->data->drv_name = pci_dev->driver->driver.name; } /** @@ -118,7 +114,6 @@ rte_eth_dev_pci_allocate(struct rte_pci_device *dev, size_t private_data_size) } eth_dev->device = &dev->device; - eth_dev->intr_handle = &dev->intr_handle; rte_eth_copy_pci_info(eth_dev, dev); return eth_dev; } @@ -134,6 +129,12 @@ rte_eth_dev_pci_release(struct rte_eth_dev *eth_dev) eth_dev->data->dev_private = NULL; + /* + * Secondary process will check the name to attach. + * Clear this field to avoid attaching a released ports. + */ + eth_dev->data->name[0] = '\0'; + eth_dev->device = NULL; eth_dev->intr_handle = NULL; } diff --git a/lib/librte_ether/rte_ethdev_vdev.h b/lib/librte_ether/rte_ethdev_vdev.h index fa2cb61e..4d2c3e2b 100644 --- a/lib/librte_ether/rte_ethdev_vdev.h +++ b/lib/librte_ether/rte_ethdev_vdev.h @@ -77,7 +77,6 @@ rte_eth_vdev_allocate(struct rte_vdev_device *dev, size_t private_data_size) eth_dev->data->kdrv = RTE_KDRV_NONE; eth_dev->data->numa_node = dev->device.numa_node; - eth_dev->data->drv_name = dev->device.driver->name; return eth_dev; } diff --git a/lib/librte_ether/rte_ether_version.map b/lib/librte_ether/rte_ether_version.map index d6726bb1..42837285 100644 --- a/lib/librte_ether/rte_ether_version.map +++ b/lib/librte_ether/rte_ether_version.map @@ -1,7 +1,6 @@ DPDK_2.2 { global: - _rte_eth_dev_callback_process; rte_eth_add_rx_callback; rte_eth_add_tx_callback; rte_eth_allmulticast_disable; @@ -10,14 +9,6 @@ DPDK_2.2 { rte_eth_dev_allocate; rte_eth_dev_allocated; rte_eth_dev_attach; - rte_eth_dev_bypass_event_show; - rte_eth_dev_bypass_event_store; - rte_eth_dev_bypass_init; - rte_eth_dev_bypass_state_set; - rte_eth_dev_bypass_state_show; - rte_eth_dev_bypass_ver_show; - rte_eth_dev_bypass_wd_reset; - rte_eth_dev_bypass_wd_timeout_show; rte_eth_dev_callback_register; rte_eth_dev_callback_unregister; rte_eth_dev_close; @@ -70,7 +61,6 @@ DPDK_2.2 { rte_eth_dev_uc_all_hash_table_set; rte_eth_dev_uc_hash_table_set; rte_eth_dev_vlan_filter; - rte_eth_dev_wd_timeout_store; rte_eth_dma_zone_reserve; rte_eth_led_off; rte_eth_led_on; @@ -151,8 +141,49 @@ DPDK_17.05 { rte_eth_dev_attach_secondary; rte_eth_find_next; + rte_eth_tx_done_cleanup; rte_eth_xstats_get_by_id; rte_eth_xstats_get_id_by_name; rte_eth_xstats_get_names_by_id; } DPDK_17.02; + +DPDK_17.08 { + global: + + _rte_eth_dev_callback_process; + rte_eth_dev_adjust_nb_rx_tx_desc; + rte_flow_copy; + rte_flow_isolate; + rte_tm_capabilities_get; + rte_tm_get_leaf_nodes; + rte_tm_hierarchy_commit; + rte_tm_level_capabilities_get; + rte_tm_mark_ip_dscp; + rte_tm_mark_ip_ecn; + rte_tm_mark_vlan_dei; + rte_tm_node_add; + rte_tm_node_capabilities_get; + rte_tm_node_cman_update; + rte_tm_node_delete; + rte_tm_node_parent_update; + rte_tm_node_resume; + rte_tm_node_shaper_update; + rte_tm_node_shared_shaper_update; + rte_tm_node_shared_wred_context_update; + rte_tm_node_stats_read; + rte_tm_node_stats_update; + rte_tm_node_suspend; + rte_tm_node_type_get; + rte_tm_node_wfq_weight_mode_update; + rte_tm_node_wred_context_update; + rte_tm_shaper_profile_add; + rte_tm_shaper_profile_delete; + rte_tm_shared_shaper_add_update; + rte_tm_shared_shaper_delete; + rte_tm_shared_wred_context_add_update; + rte_tm_shared_wred_context_delete; + rte_tm_wred_profile_add; + rte_tm_wred_profile_delete; + +} DPDK_17.05; diff --git a/lib/librte_ether/rte_flow.c b/lib/librte_ether/rte_flow.c index aaa70d68..2001fbbf 100644 --- a/lib/librte_ether/rte_flow.c +++ b/lib/librte_ether/rte_flow.c @@ -31,14 +31,81 @@ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ +#include +#include #include +#include +#include #include #include #include "rte_ethdev.h" #include "rte_flow_driver.h" #include "rte_flow.h" +/** + * Flow elements description tables. + */ +struct rte_flow_desc_data { + const char *name; + size_t size; +}; + +/** Generate flow_item[] entry. */ +#define MK_FLOW_ITEM(t, s) \ + [RTE_FLOW_ITEM_TYPE_ ## t] = { \ + .name = # t, \ + .size = s, \ + } + +/** Information about known flow pattern items. */ +static const struct rte_flow_desc_data rte_flow_desc_item[] = { + MK_FLOW_ITEM(END, 0), + MK_FLOW_ITEM(VOID, 0), + MK_FLOW_ITEM(INVERT, 0), + MK_FLOW_ITEM(ANY, sizeof(struct rte_flow_item_any)), + MK_FLOW_ITEM(PF, 0), + MK_FLOW_ITEM(VF, sizeof(struct rte_flow_item_vf)), + MK_FLOW_ITEM(PORT, sizeof(struct rte_flow_item_port)), + MK_FLOW_ITEM(RAW, sizeof(struct rte_flow_item_raw)), /* +pattern[] */ + MK_FLOW_ITEM(ETH, sizeof(struct rte_flow_item_eth)), + MK_FLOW_ITEM(VLAN, sizeof(struct rte_flow_item_vlan)), + MK_FLOW_ITEM(IPV4, sizeof(struct rte_flow_item_ipv4)), + MK_FLOW_ITEM(IPV6, sizeof(struct rte_flow_item_ipv6)), + MK_FLOW_ITEM(ICMP, sizeof(struct rte_flow_item_icmp)), + MK_FLOW_ITEM(UDP, sizeof(struct rte_flow_item_udp)), + MK_FLOW_ITEM(TCP, sizeof(struct rte_flow_item_tcp)), + MK_FLOW_ITEM(SCTP, sizeof(struct rte_flow_item_sctp)), + MK_FLOW_ITEM(VXLAN, sizeof(struct rte_flow_item_vxlan)), + MK_FLOW_ITEM(MPLS, sizeof(struct rte_flow_item_mpls)), + MK_FLOW_ITEM(GRE, sizeof(struct rte_flow_item_gre)), + MK_FLOW_ITEM(E_TAG, sizeof(struct rte_flow_item_e_tag)), + MK_FLOW_ITEM(NVGRE, sizeof(struct rte_flow_item_nvgre)), +}; + +/** Generate flow_action[] entry. */ +#define MK_FLOW_ACTION(t, s) \ + [RTE_FLOW_ACTION_TYPE_ ## t] = { \ + .name = # t, \ + .size = s, \ + } + +/** Information about known flow actions. */ +static const struct rte_flow_desc_data rte_flow_desc_action[] = { + MK_FLOW_ACTION(END, 0), + MK_FLOW_ACTION(VOID, 0), + MK_FLOW_ACTION(PASSTHRU, 0), + MK_FLOW_ACTION(MARK, sizeof(struct rte_flow_action_mark)), + MK_FLOW_ACTION(FLAG, 0), + MK_FLOW_ACTION(QUEUE, sizeof(struct rte_flow_action_queue)), + MK_FLOW_ACTION(DROP, 0), + MK_FLOW_ACTION(COUNT, 0), + MK_FLOW_ACTION(DUP, sizeof(struct rte_flow_action_dup)), + MK_FLOW_ACTION(RSS, sizeof(struct rte_flow_action_rss)), /* +queue[] */ + MK_FLOW_ACTION(PF, 0), + MK_FLOW_ACTION(VF, sizeof(struct rte_flow_action_vf)), +}; + /* Get generic flow operations structure from a port. */ const struct rte_flow_ops * rte_flow_ops_get(uint8_t port_id, struct rte_flow_error *error) @@ -157,3 +224,185 @@ rte_flow_query(uint8_t port_id, RTE_FLOW_ERROR_TYPE_UNSPECIFIED, NULL, rte_strerror(ENOSYS)); } + +/* Restrict ingress traffic to the defined flow rules. */ +int +rte_flow_isolate(uint8_t port_id, + int set, + struct rte_flow_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + const struct rte_flow_ops *ops = rte_flow_ops_get(port_id, error); + + if (!ops) + return -rte_errno; + if (likely(!!ops->isolate)) + return ops->isolate(dev, set, error); + return -rte_flow_error_set(error, ENOSYS, + RTE_FLOW_ERROR_TYPE_UNSPECIFIED, + NULL, rte_strerror(ENOSYS)); +} + +/** Compute storage space needed by item specification. */ +static void +flow_item_spec_size(const struct rte_flow_item *item, + size_t *size, size_t *pad) +{ + if (!item->spec) { + *size = 0; + goto empty; + } + switch (item->type) { + union { + const struct rte_flow_item_raw *raw; + } spec; + + /* Not a fall-through */ + case RTE_FLOW_ITEM_TYPE_RAW: + spec.raw = item->spec; + *size = offsetof(struct rte_flow_item_raw, pattern) + + spec.raw->length * sizeof(*spec.raw->pattern); + break; + default: + *size = rte_flow_desc_item[item->type].size; + break; + } +empty: + *pad = RTE_ALIGN_CEIL(*size, sizeof(double)) - *size; +} + +/** Compute storage space needed by action configuration. */ +static void +flow_action_conf_size(const struct rte_flow_action *action, + size_t *size, size_t *pad) +{ + if (!action->conf) { + *size = 0; + goto empty; + } + switch (action->type) { + union { + const struct rte_flow_action_rss *rss; + } conf; + + /* Not a fall-through. */ + case RTE_FLOW_ACTION_TYPE_RSS: + conf.rss = action->conf; + *size = offsetof(struct rte_flow_action_rss, queue) + + conf.rss->num * sizeof(*conf.rss->queue); + break; + default: + *size = rte_flow_desc_action[action->type].size; + break; + } +empty: + *pad = RTE_ALIGN_CEIL(*size, sizeof(double)) - *size; +} + +/** Store a full rte_flow description. */ +size_t +rte_flow_copy(struct rte_flow_desc *desc, size_t len, + const struct rte_flow_attr *attr, + const struct rte_flow_item *items, + const struct rte_flow_action *actions) +{ + struct rte_flow_desc *fd = NULL; + size_t tmp; + size_t pad; + size_t off1 = 0; + size_t off2 = 0; + size_t size = 0; + +store: + if (items) { + const struct rte_flow_item *item; + + item = items; + if (fd) + fd->items = (void *)&fd->data[off1]; + do { + struct rte_flow_item *dst = NULL; + + if ((size_t)item->type >= + RTE_DIM(rte_flow_desc_item) || + !rte_flow_desc_item[item->type].name) { + rte_errno = ENOTSUP; + return 0; + } + if (fd) + dst = memcpy(fd->data + off1, item, + sizeof(*item)); + off1 += sizeof(*item); + flow_item_spec_size(item, &tmp, &pad); + if (item->spec) { + if (fd) + dst->spec = memcpy(fd->data + off2, + item->spec, tmp); + off2 += tmp + pad; + } + if (item->last) { + if (fd) + dst->last = memcpy(fd->data + off2, + item->last, tmp); + off2 += tmp + pad; + } + if (item->mask) { + if (fd) + dst->mask = memcpy(fd->data + off2, + item->mask, tmp); + off2 += tmp + pad; + } + off2 = RTE_ALIGN_CEIL(off2, sizeof(double)); + } while ((item++)->type != RTE_FLOW_ITEM_TYPE_END); + off1 = RTE_ALIGN_CEIL(off1, sizeof(double)); + } + if (actions) { + const struct rte_flow_action *action; + + action = actions; + if (fd) + fd->actions = (void *)&fd->data[off1]; + do { + struct rte_flow_action *dst = NULL; + + if ((size_t)action->type >= + RTE_DIM(rte_flow_desc_action) || + !rte_flow_desc_action[action->type].name) { + rte_errno = ENOTSUP; + return 0; + } + if (fd) + dst = memcpy(fd->data + off1, action, + sizeof(*action)); + off1 += sizeof(*action); + flow_action_conf_size(action, &tmp, &pad); + if (action->conf) { + if (fd) + dst->conf = memcpy(fd->data + off2, + action->conf, tmp); + off2 += tmp + pad; + } + off2 = RTE_ALIGN_CEIL(off2, sizeof(double)); + } while ((action++)->type != RTE_FLOW_ACTION_TYPE_END); + } + if (fd != NULL) + return size; + off1 = RTE_ALIGN_CEIL(off1, sizeof(double)); + tmp = RTE_ALIGN_CEIL(offsetof(struct rte_flow_desc, data), + sizeof(double)); + size = tmp + off1 + off2; + if (size > len) + return size; + fd = desc; + if (fd != NULL) { + *fd = (const struct rte_flow_desc) { + .size = size, + .attr = *attr, + }; + tmp -= offsetof(struct rte_flow_desc, data); + off2 = tmp + off1; + off1 = tmp; + goto store; + } + return 0; +} diff --git a/lib/librte_ether/rte_flow.h b/lib/librte_ether/rte_flow.h index c47edbc9..bba6169f 100644 --- a/lib/librte_ether/rte_flow.h +++ b/lib/librte_ether/rte_flow.h @@ -297,6 +297,18 @@ enum rte_flow_item_type { * See struct rte_flow_item_gre. */ RTE_FLOW_ITEM_TYPE_GRE, + + /** + * [META] + * + * Fuzzy pattern match, expect faster than default. + * + * This is for device that support fuzzy matching option. + * Usually a fuzzy matching is fast but the cost is accuracy. + * + * See struct rte_flow_item_fuzzy. + */ + RTE_FLOW_ITEM_TYPE_FUZZY, }; /** @@ -429,7 +441,7 @@ static const struct rte_flow_item_raw rte_flow_item_raw_mask = { struct rte_flow_item_eth { struct ether_addr dst; /**< Destination MAC. */ struct ether_addr src; /**< Source MAC. */ - uint16_t type; /**< EtherType. */ + rte_be16_t type; /**< EtherType. */ }; /** Default mask for RTE_FLOW_ITEM_TYPE_ETH. */ @@ -437,7 +449,7 @@ struct rte_flow_item_eth { static const struct rte_flow_item_eth rte_flow_item_eth_mask = { .dst.addr_bytes = "\xff\xff\xff\xff\xff\xff", .src.addr_bytes = "\xff\xff\xff\xff\xff\xff", - .type = 0x0000, + .type = RTE_BE16(0x0000), }; #endif @@ -450,15 +462,15 @@ static const struct rte_flow_item_eth rte_flow_item_eth_mask = { * RTE_FLOW_ITEM_TYPE_VLAN. */ struct rte_flow_item_vlan { - uint16_t tpid; /**< Tag protocol identifier. */ - uint16_t tci; /**< Tag control information. */ + rte_be16_t tpid; /**< Tag protocol identifier. */ + rte_be16_t tci; /**< Tag control information. */ }; /** Default mask for RTE_FLOW_ITEM_TYPE_VLAN. */ #ifndef __cplusplus static const struct rte_flow_item_vlan rte_flow_item_vlan_mask = { - .tpid = 0x0000, - .tci = 0xffff, + .tpid = RTE_BE16(0x0000), + .tci = RTE_BE16(0xffff), }; #endif @@ -477,8 +489,8 @@ struct rte_flow_item_ipv4 { #ifndef __cplusplus static const struct rte_flow_item_ipv4 rte_flow_item_ipv4_mask = { .hdr = { - .src_addr = 0xffffffff, - .dst_addr = 0xffffffff, + .src_addr = RTE_BE32(0xffffffff), + .dst_addr = RTE_BE32(0xffffffff), }, }; #endif @@ -540,8 +552,8 @@ struct rte_flow_item_udp { #ifndef __cplusplus static const struct rte_flow_item_udp rte_flow_item_udp_mask = { .hdr = { - .src_port = 0xffff, - .dst_port = 0xffff, + .src_port = RTE_BE16(0xffff), + .dst_port = RTE_BE16(0xffff), }, }; #endif @@ -559,8 +571,8 @@ struct rte_flow_item_tcp { #ifndef __cplusplus static const struct rte_flow_item_tcp rte_flow_item_tcp_mask = { .hdr = { - .src_port = 0xffff, - .dst_port = 0xffff, + .src_port = RTE_BE16(0xffff), + .dst_port = RTE_BE16(0xffff), }, }; #endif @@ -578,8 +590,8 @@ struct rte_flow_item_sctp { #ifndef __cplusplus static const struct rte_flow_item_sctp rte_flow_item_sctp_mask = { .hdr = { - .src_port = 0xffff, - .dst_port = 0xffff, + .src_port = RTE_BE16(0xffff), + .dst_port = RTE_BE16(0xffff), }, }; #endif @@ -609,14 +621,14 @@ static const struct rte_flow_item_vxlan rte_flow_item_vxlan_mask = { * Matches a E-tag header. */ struct rte_flow_item_e_tag { - uint16_t tpid; /**< Tag protocol identifier (0x893F). */ + rte_be16_t tpid; /**< Tag protocol identifier (0x893F). */ /** * E-Tag control information (E-TCI). * E-PCP (3b), E-DEI (1b), ingress E-CID base (12b). */ - uint16_t epcp_edei_in_ecid_b; + rte_be16_t epcp_edei_in_ecid_b; /** Reserved (2b), GRP (2b), E-CID base (12b). */ - uint16_t rsvd_grp_ecid_b; + rte_be16_t rsvd_grp_ecid_b; uint8_t in_ecid_e; /**< Ingress E-CID ext. */ uint8_t ecid_e; /**< E-CID ext. */ }; @@ -624,13 +636,7 @@ struct rte_flow_item_e_tag { /** Default mask for RTE_FLOW_ITEM_TYPE_E_TAG. */ #ifndef __cplusplus static const struct rte_flow_item_e_tag rte_flow_item_e_tag_mask = { -#if RTE_BYTE_ORDER == RTE_BIG_ENDIAN - .rsvd_grp_ecid_b = 0x3fff, -#elif RTE_BYTE_ORDER == RTE_LITTLE_ENDIAN - .rsvd_grp_ecid_b = 0xff3f, -#else -#error Unsupported endianness. -#endif + .rsvd_grp_ecid_b = RTE_BE16(0x3fff), }; #endif @@ -646,8 +652,8 @@ struct rte_flow_item_nvgre { * * c_k_s_rsvd0_ver must have value 0x2000 according to RFC 7637. */ - uint16_t c_k_s_rsvd0_ver; - uint16_t protocol; /**< Protocol type (0x6558). */ + rte_be16_t c_k_s_rsvd0_ver; + rte_be16_t protocol; /**< Protocol type (0x6558). */ uint8_t tni[3]; /**< Virtual subnet ID. */ uint8_t flow_id; /**< Flow ID. */ }; @@ -689,14 +695,42 @@ struct rte_flow_item_gre { * Checksum (1b), reserved 0 (12b), version (3b). * Refer to RFC 2784. */ - uint16_t c_rsvd0_ver; - uint16_t protocol; /**< Protocol type. */ + rte_be16_t c_rsvd0_ver; + rte_be16_t protocol; /**< Protocol type. */ }; /** Default mask for RTE_FLOW_ITEM_TYPE_GRE. */ #ifndef __cplusplus static const struct rte_flow_item_gre rte_flow_item_gre_mask = { - .protocol = 0xffff, + .protocol = RTE_BE16(0xffff), +}; +#endif + +/** + * RTE_FLOW_ITEM_TYPE_FUZZY + * + * Fuzzy pattern match, expect faster than default. + * + * This is for device that support fuzzy match option. + * Usually a fuzzy match is fast but the cost is accuracy. + * i.e. Signature Match only match pattern's hash value, but it is + * possible two different patterns have the same hash value. + * + * Matching accuracy level can be configure by threshold. + * Driver can divide the range of threshold and map to different + * accuracy levels that device support. + * + * Threshold 0 means perfect match (no fuzziness), while threshold + * 0xffffffff means fuzziest match. + */ +struct rte_flow_item_fuzzy { + uint32_t thresh; /**< Accuracy threshold. */ +}; + +/** Default mask for RTE_FLOW_ITEM_TYPE_FUZZY. */ +#ifndef __cplusplus +static const struct rte_flow_item_fuzzy rte_flow_item_fuzzy_mask = { + .thresh = 0xffffffff, }; #endif @@ -1191,6 +1225,90 @@ rte_flow_query(uint8_t port_id, void *data, struct rte_flow_error *error); +/** + * Restrict ingress traffic to the defined flow rules. + * + * Isolated mode guarantees that all ingress traffic comes from defined flow + * rules only (current and future). + * + * Besides making ingress more deterministic, it allows PMDs to safely reuse + * resources otherwise assigned to handle the remaining traffic, such as + * global RSS configuration settings, VLAN filters, MAC address entries, + * legacy filter API rules and so on in order to expand the set of possible + * flow rule types. + * + * Calling this function as soon as possible after device initialization, + * ideally before the first call to rte_eth_dev_configure(), is recommended + * to avoid possible failures due to conflicting settings. + * + * Once effective, leaving isolated mode may not be possible depending on + * PMD implementation. + * + * Additionally, the following functionality has no effect on the underlying + * port and may return errors such as ENOTSUP ("not supported"): + * + * - Toggling promiscuous mode. + * - Toggling allmulticast mode. + * - Configuring MAC addresses. + * - Configuring multicast addresses. + * - Configuring VLAN filters. + * - Configuring Rx filters through the legacy API (e.g. FDIR). + * - Configuring global RSS settings. + * + * @param port_id + * Port identifier of Ethernet device. + * @param set + * Nonzero to enter isolated mode, attempt to leave it otherwise. + * @param[out] error + * Perform verbose error reporting if not NULL. PMDs initialize this + * structure in case of error only. + * + * @return + * 0 on success, a negative errno value otherwise and rte_errno is set. + */ +int +rte_flow_isolate(uint8_t port_id, int set, struct rte_flow_error *error); + +/** + * Generic flow representation. + * + * This form is sufficient to describe an rte_flow independently from any + * PMD implementation and allows for replayability and identification. + */ +struct rte_flow_desc { + size_t size; /**< Allocated space including data[]. */ + struct rte_flow_attr attr; /**< Attributes. */ + struct rte_flow_item *items; /**< Items. */ + struct rte_flow_action *actions; /**< Actions. */ + uint8_t data[]; /**< Storage for items/actions. */ +}; + +/** + * Copy an rte_flow rule description. + * + * @param[in] fd + * Flow rule description. + * @param[in] len + * Total size of allocated data for the flow description. + * @param[in] attr + * Flow rule attributes. + * @param[in] items + * Pattern specification (list terminated by the END pattern item). + * @param[in] actions + * Associated actions (list terminated by the END action). + * + * @return + * If len is greater or equal to the size of the flow, the total size of the + * flow description and its data. + * If len is lower than the size of the flow, the number of bytes that would + * have been written to desc had it been sufficient. Nothing is written. + */ +size_t +rte_flow_copy(struct rte_flow_desc *fd, size_t len, + const struct rte_flow_attr *attr, + const struct rte_flow_item *items, + const struct rte_flow_action *actions); + #ifdef __cplusplus } #endif diff --git a/lib/librte_ether/rte_flow_driver.h b/lib/librte_ether/rte_flow_driver.h index da5749d5..4d95391d 100644 --- a/lib/librte_ether/rte_flow_driver.h +++ b/lib/librte_ether/rte_flow_driver.h @@ -120,6 +120,11 @@ struct rte_flow_ops { enum rte_flow_action_type, void *, struct rte_flow_error *); + /** See rte_flow_isolate(). */ + int (*isolate) + (struct rte_eth_dev *, + int, + struct rte_flow_error *); }; /** diff --git a/lib/librte_ether/rte_tm.c b/lib/librte_ether/rte_tm.c new file mode 100644 index 00000000..71679650 --- /dev/null +++ b/lib/librte_ether/rte_tm.c @@ -0,0 +1,438 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2017 Intel Corporation. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#include + +#include +#include "rte_ethdev.h" +#include "rte_tm_driver.h" +#include "rte_tm.h" + +/* Get generic traffic manager operations structure from a port. */ +const struct rte_tm_ops * +rte_tm_ops_get(uint8_t port_id, struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + const struct rte_tm_ops *ops; + + if (!rte_eth_dev_is_valid_port(port_id)) { + rte_tm_error_set(error, + ENODEV, + RTE_TM_ERROR_TYPE_UNSPECIFIED, + NULL, + rte_strerror(ENODEV)); + return NULL; + } + + if ((dev->dev_ops->tm_ops_get == NULL) || + (dev->dev_ops->tm_ops_get(dev, &ops) != 0) || + (ops == NULL)) { + rte_tm_error_set(error, + ENOSYS, + RTE_TM_ERROR_TYPE_UNSPECIFIED, + NULL, + rte_strerror(ENOSYS)); + return NULL; + } + + return ops; +} + +#define RTE_TM_FUNC(port_id, func) \ +({ \ + const struct rte_tm_ops *ops = \ + rte_tm_ops_get(port_id, error); \ + if (ops == NULL) \ + return -rte_errno; \ + \ + if (ops->func == NULL) \ + return -rte_tm_error_set(error, \ + ENOSYS, \ + RTE_TM_ERROR_TYPE_UNSPECIFIED, \ + NULL, \ + rte_strerror(ENOSYS)); \ + \ + ops->func; \ +}) + +/* Get number of leaf nodes */ +int +rte_tm_get_number_of_leaf_nodes(uint8_t port_id, + uint32_t *n_leaf_nodes, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + const struct rte_tm_ops *ops = + rte_tm_ops_get(port_id, error); + + if (ops == NULL) + return -rte_errno; + + if (n_leaf_nodes == NULL) { + rte_tm_error_set(error, + EINVAL, + RTE_TM_ERROR_TYPE_UNSPECIFIED, + NULL, + rte_strerror(EINVAL)); + return -rte_errno; + } + + *n_leaf_nodes = dev->data->nb_tx_queues; + return 0; +} + +/* Check node type (leaf or non-leaf) */ +int +rte_tm_node_type_get(uint8_t port_id, + uint32_t node_id, + int *is_leaf, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_type_get)(dev, + node_id, is_leaf, error); +} + +/* Get capabilities */ +int rte_tm_capabilities_get(uint8_t port_id, + struct rte_tm_capabilities *cap, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, capabilities_get)(dev, + cap, error); +} + +/* Get level capabilities */ +int rte_tm_level_capabilities_get(uint8_t port_id, + uint32_t level_id, + struct rte_tm_level_capabilities *cap, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, level_capabilities_get)(dev, + level_id, cap, error); +} + +/* Get node capabilities */ +int rte_tm_node_capabilities_get(uint8_t port_id, + uint32_t node_id, + struct rte_tm_node_capabilities *cap, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_capabilities_get)(dev, + node_id, cap, error); +} + +/* Add WRED profile */ +int rte_tm_wred_profile_add(uint8_t port_id, + uint32_t wred_profile_id, + struct rte_tm_wred_params *profile, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, wred_profile_add)(dev, + wred_profile_id, profile, error); +} + +/* Delete WRED profile */ +int rte_tm_wred_profile_delete(uint8_t port_id, + uint32_t wred_profile_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, wred_profile_delete)(dev, + wred_profile_id, error); +} + +/* Add/update shared WRED context */ +int rte_tm_shared_wred_context_add_update(uint8_t port_id, + uint32_t shared_wred_context_id, + uint32_t wred_profile_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, shared_wred_context_add_update)(dev, + shared_wred_context_id, wred_profile_id, error); +} + +/* Delete shared WRED context */ +int rte_tm_shared_wred_context_delete(uint8_t port_id, + uint32_t shared_wred_context_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, shared_wred_context_delete)(dev, + shared_wred_context_id, error); +} + +/* Add shaper profile */ +int rte_tm_shaper_profile_add(uint8_t port_id, + uint32_t shaper_profile_id, + struct rte_tm_shaper_params *profile, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, shaper_profile_add)(dev, + shaper_profile_id, profile, error); +} + +/* Delete WRED profile */ +int rte_tm_shaper_profile_delete(uint8_t port_id, + uint32_t shaper_profile_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, shaper_profile_delete)(dev, + shaper_profile_id, error); +} + +/* Add shared shaper */ +int rte_tm_shared_shaper_add_update(uint8_t port_id, + uint32_t shared_shaper_id, + uint32_t shaper_profile_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, shared_shaper_add_update)(dev, + shared_shaper_id, shaper_profile_id, error); +} + +/* Delete shared shaper */ +int rte_tm_shared_shaper_delete(uint8_t port_id, + uint32_t shared_shaper_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, shared_shaper_delete)(dev, + shared_shaper_id, error); +} + +/* Add node to port traffic manager hierarchy */ +int rte_tm_node_add(uint8_t port_id, + uint32_t node_id, + uint32_t parent_node_id, + uint32_t priority, + uint32_t weight, + uint32_t level_id, + struct rte_tm_node_params *params, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_add)(dev, + node_id, parent_node_id, priority, weight, level_id, + params, error); +} + +/* Delete node from traffic manager hierarchy */ +int rte_tm_node_delete(uint8_t port_id, + uint32_t node_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_delete)(dev, + node_id, error); +} + +/* Suspend node */ +int rte_tm_node_suspend(uint8_t port_id, + uint32_t node_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_suspend)(dev, + node_id, error); +} + +/* Resume node */ +int rte_tm_node_resume(uint8_t port_id, + uint32_t node_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_resume)(dev, + node_id, error); +} + +/* Commit the initial port traffic manager hierarchy */ +int rte_tm_hierarchy_commit(uint8_t port_id, + int clear_on_fail, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, hierarchy_commit)(dev, + clear_on_fail, error); +} + +/* Update node parent */ +int rte_tm_node_parent_update(uint8_t port_id, + uint32_t node_id, + uint32_t parent_node_id, + uint32_t priority, + uint32_t weight, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_parent_update)(dev, + node_id, parent_node_id, priority, weight, error); +} + +/* Update node private shaper */ +int rte_tm_node_shaper_update(uint8_t port_id, + uint32_t node_id, + uint32_t shaper_profile_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_shaper_update)(dev, + node_id, shaper_profile_id, error); +} + +/* Update node shared shapers */ +int rte_tm_node_shared_shaper_update(uint8_t port_id, + uint32_t node_id, + uint32_t shared_shaper_id, + int add, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_shared_shaper_update)(dev, + node_id, shared_shaper_id, add, error); +} + +/* Update node stats */ +int rte_tm_node_stats_update(uint8_t port_id, + uint32_t node_id, + uint64_t stats_mask, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_stats_update)(dev, + node_id, stats_mask, error); +} + +/* Update WFQ weight mode */ +int rte_tm_node_wfq_weight_mode_update(uint8_t port_id, + uint32_t node_id, + int *wfq_weight_mode, + uint32_t n_sp_priorities, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_wfq_weight_mode_update)(dev, + node_id, wfq_weight_mode, n_sp_priorities, error); +} + +/* Update node congestion management mode */ +int rte_tm_node_cman_update(uint8_t port_id, + uint32_t node_id, + enum rte_tm_cman_mode cman, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_cman_update)(dev, + node_id, cman, error); +} + +/* Update node private WRED context */ +int rte_tm_node_wred_context_update(uint8_t port_id, + uint32_t node_id, + uint32_t wred_profile_id, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_wred_context_update)(dev, + node_id, wred_profile_id, error); +} + +/* Update node shared WRED context */ +int rte_tm_node_shared_wred_context_update(uint8_t port_id, + uint32_t node_id, + uint32_t shared_wred_context_id, + int add, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_shared_wred_context_update)(dev, + node_id, shared_wred_context_id, add, error); +} + +/* Read and/or clear stats counters for specific node */ +int rte_tm_node_stats_read(uint8_t port_id, + uint32_t node_id, + struct rte_tm_node_stats *stats, + uint64_t *stats_mask, + int clear, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, node_stats_read)(dev, + node_id, stats, stats_mask, clear, error); +} + +/* Packet marking - VLAN DEI */ +int rte_tm_mark_vlan_dei(uint8_t port_id, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, mark_vlan_dei)(dev, + mark_green, mark_yellow, mark_red, error); +} + +/* Packet marking - IPv4/IPv6 ECN */ +int rte_tm_mark_ip_ecn(uint8_t port_id, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, mark_ip_ecn)(dev, + mark_green, mark_yellow, mark_red, error); +} + +/* Packet marking - IPv4/IPv6 DSCP */ +int rte_tm_mark_ip_dscp(uint8_t port_id, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error) +{ + struct rte_eth_dev *dev = &rte_eth_devices[port_id]; + return RTE_TM_FUNC(port_id, mark_ip_dscp)(dev, + mark_green, mark_yellow, mark_red, error); +} diff --git a/lib/librte_ether/rte_tm.h b/lib/librte_ether/rte_tm.h new file mode 100644 index 00000000..ebbfa1ee --- /dev/null +++ b/lib/librte_ether/rte_tm.h @@ -0,0 +1,1912 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2017 Intel Corporation. + * Copyright(c) 2017 Cavium. + * Copyright(c) 2017 NXP. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#ifndef __INCLUDE_RTE_TM_H__ +#define __INCLUDE_RTE_TM_H__ + +/** + * @file + * RTE Generic Traffic Manager API + * + * This interface provides the ability to configure the traffic manager in a + * generic way. It includes features such as: hierarchical scheduling, + * traffic shaping, congestion management, packet marking, etc. + * + * @warning + * @b EXPERIMENTAL: this API may change without prior notice + */ + +#include + +#include + +#ifdef __cplusplus +extern "C" { +#endif + +/** + * Ethernet framing overhead. + * + * Overhead fields per Ethernet frame: + * 1. Preamble: 7 bytes; + * 2. Start of Frame Delimiter (SFD): 1 byte; + * 3. Inter-Frame Gap (IFG): 12 bytes. + * + * One of the typical values for the *pkt_length_adjust* field of the shaper + * profile. + * + * @see struct rte_tm_shaper_params + */ +#define RTE_TM_ETH_FRAMING_OVERHEAD 20 + +/** + * Ethernet framing overhead including the Frame Check Sequence (FCS) field. + * Useful when FCS is generated and added at the end of the Ethernet frame on + * TX side without any SW intervention. + * + * One of the typical values for the pkt_length_adjust field of the shaper + * profile. + * + * @see struct rte_tm_shaper_params + */ +#define RTE_TM_ETH_FRAMING_OVERHEAD_FCS 24 + +/** + * Invalid WRED profile ID. + * + * @see struct rte_tm_node_params + * @see rte_tm_node_add() + * @see rte_tm_node_wred_context_update() + */ +#define RTE_TM_WRED_PROFILE_ID_NONE UINT32_MAX + +/** + *Invalid shaper profile ID. + * + * @see struct rte_tm_node_params + * @see rte_tm_node_add() + * @see rte_tm_node_shaper_update() + */ +#define RTE_TM_SHAPER_PROFILE_ID_NONE UINT32_MAX + +/** + * Node ID for the parent of the root node. + * + * @see rte_tm_node_add() + */ +#define RTE_TM_NODE_ID_NULL UINT32_MAX + +/** + * Node level ID used to disable level ID checking. + * + * @see rte_tm_node_add() + */ +#define RTE_TM_NODE_LEVEL_ID_ANY UINT32_MAX + +/** + * Color + */ +enum rte_tm_color { + RTE_TM_GREEN = 0, /**< Green */ + RTE_TM_YELLOW, /**< Yellow */ + RTE_TM_RED, /**< Red */ + RTE_TM_COLORS /**< Number of colors */ +}; + +/** + * Node statistics counter type + */ +enum rte_tm_stats_type { + /** Number of packets scheduled from current node. */ + RTE_TM_STATS_N_PKTS = 1 << 0, + + /** Number of bytes scheduled from current node. */ + RTE_TM_STATS_N_BYTES = 1 << 1, + + /** Number of green packets dropped by current leaf node. */ + RTE_TM_STATS_N_PKTS_GREEN_DROPPED = 1 << 2, + + /** Number of yellow packets dropped by current leaf node. */ + RTE_TM_STATS_N_PKTS_YELLOW_DROPPED = 1 << 3, + + /** Number of red packets dropped by current leaf node. */ + RTE_TM_STATS_N_PKTS_RED_DROPPED = 1 << 4, + + /** Number of green bytes dropped by current leaf node. */ + RTE_TM_STATS_N_BYTES_GREEN_DROPPED = 1 << 5, + + /** Number of yellow bytes dropped by current leaf node. */ + RTE_TM_STATS_N_BYTES_YELLOW_DROPPED = 1 << 6, + + /** Number of red bytes dropped by current leaf node. */ + RTE_TM_STATS_N_BYTES_RED_DROPPED = 1 << 7, + + /** Number of packets currently waiting in the packet queue of current + * leaf node. + */ + RTE_TM_STATS_N_PKTS_QUEUED = 1 << 8, + + /** Number of bytes currently waiting in the packet queue of current + * leaf node. + */ + RTE_TM_STATS_N_BYTES_QUEUED = 1 << 9, +}; + +/** + * Node statistics counters + */ +struct rte_tm_node_stats { + /** Number of packets scheduled from current node. */ + uint64_t n_pkts; + + /** Number of bytes scheduled from current node. */ + uint64_t n_bytes; + + /** Statistics counters for leaf nodes only. */ + struct { + /** Number of packets dropped by current leaf node per each + * color. + */ + uint64_t n_pkts_dropped[RTE_TM_COLORS]; + + /** Number of bytes dropped by current leaf node per each + * color. + */ + uint64_t n_bytes_dropped[RTE_TM_COLORS]; + + /** Number of packets currently waiting in the packet queue of + * current leaf node. + */ + uint64_t n_pkts_queued; + + /** Number of bytes currently waiting in the packet queue of + * current leaf node. + */ + uint64_t n_bytes_queued; + } leaf; +}; + +/** + * Traffic manager dynamic updates + */ +enum rte_tm_dynamic_update_type { + /** Dynamic parent node update. The new parent node is located on same + * hierarchy level as the former parent node. Consequently, the node + * whose parent is changed preserves its hierarchy level. + */ + RTE_TM_UPDATE_NODE_PARENT_KEEP_LEVEL = 1 << 0, + + /** Dynamic parent node update. The new parent node is located on + * different hierarchy level than the former parent node. Consequently, + * the node whose parent is changed also changes its hierarchy level. + */ + RTE_TM_UPDATE_NODE_PARENT_CHANGE_LEVEL = 1 << 1, + + /** Dynamic node add/delete. */ + RTE_TM_UPDATE_NODE_ADD_DELETE = 1 << 2, + + /** Suspend/resume nodes. */ + RTE_TM_UPDATE_NODE_SUSPEND_RESUME = 1 << 3, + + /** Dynamic switch between byte-based and packet-based WFQ weights. */ + RTE_TM_UPDATE_NODE_WFQ_WEIGHT_MODE = 1 << 4, + + /** Dynamic update on number of SP priorities. */ + RTE_TM_UPDATE_NODE_N_SP_PRIORITIES = 1 << 5, + + /** Dynamic update of congestion management mode for leaf nodes. */ + RTE_TM_UPDATE_NODE_CMAN = 1 << 6, + + /** Dynamic update of the set of enabled stats counter types. */ + RTE_TM_UPDATE_NODE_STATS = 1 << 7, +}; + +/** + * Traffic manager capabilities + */ +struct rte_tm_capabilities { + /** Maximum number of nodes. */ + uint32_t n_nodes_max; + + /** Maximum number of levels (i.e. number of nodes connecting the root + * node with any leaf node, including the root and the leaf). + */ + uint32_t n_levels_max; + + /** When non-zero, this flag indicates that all the non-leaf nodes + * (with the exception of the root node) have identical capability set. + */ + int non_leaf_nodes_identical; + + /** When non-zero, this flag indicates that all the leaf nodes have + * identical capability set. + */ + int leaf_nodes_identical; + + /** Maximum number of shapers, either private or shared. In case the + * implementation does not share any resources between private and + * shared shapers, it is typically equal to the sum of + * *shaper_private_n_max* and *shaper_shared_n_max*. The + * value of zero indicates that traffic shaping is not supported. + */ + uint32_t shaper_n_max; + + /** Maximum number of private shapers. Indicates the maximum number of + * nodes that can concurrently have their private shaper enabled. The + * value of zero indicates that private shapers are not supported. + */ + uint32_t shaper_private_n_max; + + /** Maximum number of private shapers that support dual rate shaping. + * Indicates the maximum number of nodes that can concurrently have + * their private shaper enabled with dual rate support. Only valid when + * private shapers are supported. The value of zero indicates that dual + * rate shaping is not available for private shapers. The maximum value + * is *shaper_private_n_max*. + */ + int shaper_private_dual_rate_n_max; + + /** Minimum committed/peak rate (bytes per second) for any private + * shaper. Valid only when private shapers are supported. + */ + uint64_t shaper_private_rate_min; + + /** Maximum committed/peak rate (bytes per second) for any private + * shaper. Valid only when private shapers are supported. + */ + uint64_t shaper_private_rate_max; + + /** Maximum number of shared shapers. The value of zero indicates that + * shared shapers are not supported. + */ + uint32_t shaper_shared_n_max; + + /** Maximum number of nodes that can share the same shared shaper. + * Only valid when shared shapers are supported. + */ + uint32_t shaper_shared_n_nodes_per_shaper_max; + + /** Maximum number of shared shapers a node can be part of. This + * parameter indicates that there is at least one node that can be + * configured with this many shared shapers, which might not be true for + * all the nodes. Only valid when shared shapers are supported, in which + * case it ranges from 1 to *shaper_shared_n_max*. + */ + uint32_t shaper_shared_n_shapers_per_node_max; + + /** Maximum number of shared shapers that can be configured with dual + * rate shaping. The value of zero indicates that dual rate shaping + * support is not available for shared shapers. + */ + uint32_t shaper_shared_dual_rate_n_max; + + /** Minimum committed/peak rate (bytes per second) for any shared + * shaper. Only valid when shared shapers are supported. + */ + uint64_t shaper_shared_rate_min; + + /** Maximum committed/peak rate (bytes per second) for any shared + * shaper. Only valid when shared shapers are supported. + */ + uint64_t shaper_shared_rate_max; + + /** Minimum value allowed for packet length adjustment for any private + * or shared shaper. + */ + int shaper_pkt_length_adjust_min; + + /** Maximum value allowed for packet length adjustment for any private + * or shared shaper. + */ + int shaper_pkt_length_adjust_max; + + /** Maximum number of children nodes. This parameter indicates that + * there is at least one non-leaf node that can be configured with this + * many children nodes, which might not be true for all the non-leaf + * nodes. + */ + uint32_t sched_n_children_max; + + /** Maximum number of supported priority levels. This parameter + * indicates that there is at least one non-leaf node that can be + * configured with this many priority levels for managing its children + * nodes, which might not be true for all the non-leaf nodes. The value + * of zero is invalid. The value of 1 indicates that only priority 0 is + * supported, which essentially means that Strict Priority (SP) + * algorithm is not supported. + */ + uint32_t sched_sp_n_priorities_max; + + /** Maximum number of sibling nodes that can have the same priority at + * any given time, i.e. maximum size of the WFQ sibling node group. This + * parameter indicates there is at least one non-leaf node that meets + * this condition, which might not be true for all the non-leaf nodes. + * The value of zero is invalid. The value of 1 indicates that WFQ + * algorithm is not supported. The maximum value is + * *sched_n_children_max*. + */ + uint32_t sched_wfq_n_children_per_group_max; + + /** Maximum number of priority levels that can have more than one child + * node at any given time, i.e. maximum number of WFQ sibling node + * groups that have two or more members. This parameter indicates there + * is at least one non-leaf node that meets this condition, which might + * not be true for all the non-leaf nodes. The value of zero states that + * WFQ algorithm is not supported. The value of 1 indicates that + * (*sched_sp_n_priorities_max* - 1) priority levels have at most one + * child node, so there can be only one priority level with two or + * more sibling nodes making up a WFQ group. The maximum value is: + * min(floor(*sched_n_children_max* / 2), *sched_sp_n_priorities_max*). + */ + uint32_t sched_wfq_n_groups_max; + + /** Maximum WFQ weight. The value of 1 indicates that all sibling nodes + * with same priority have the same WFQ weight, so WFQ is reduced to FQ. + */ + uint32_t sched_wfq_weight_max; + + /** Head drop algorithm support. When non-zero, this parameter + * indicates that there is at least one leaf node that supports the head + * drop algorithm, which might not be true for all the leaf nodes. + */ + int cman_head_drop_supported; + + /** Maximum number of WRED contexts, either private or shared. In case + * the implementation does not share any resources between private and + * shared WRED contexts, it is typically equal to the sum of + * *cman_wred_context_private_n_max* and + * *cman_wred_context_shared_n_max*. The value of zero indicates that + * WRED is not supported. + */ + uint32_t cman_wred_context_n_max; + + /** Maximum number of private WRED contexts. Indicates the maximum + * number of leaf nodes that can concurrently have their private WRED + * context enabled. The value of zero indicates that private WRED + * contexts are not supported. + */ + uint32_t cman_wred_context_private_n_max; + + /** Maximum number of shared WRED contexts. The value of zero + * indicates that shared WRED contexts are not supported. + */ + uint32_t cman_wred_context_shared_n_max; + + /** Maximum number of leaf nodes that can share the same WRED context. + * Only valid when shared WRED contexts are supported. + */ + uint32_t cman_wred_context_shared_n_nodes_per_context_max; + + /** Maximum number of shared WRED contexts a leaf node can be part of. + * This parameter indicates that there is at least one leaf node that + * can be configured with this many shared WRED contexts, which might + * not be true for all the leaf nodes. Only valid when shared WRED + * contexts are supported, in which case it ranges from 1 to + * *cman_wred_context_shared_n_max*. + */ + uint32_t cman_wred_context_shared_n_contexts_per_node_max; + + /** Support for VLAN DEI packet marking (per color). */ + int mark_vlan_dei_supported[RTE_TM_COLORS]; + + /** Support for IPv4/IPv6 ECN marking of TCP packets (per color). */ + int mark_ip_ecn_tcp_supported[RTE_TM_COLORS]; + + /** Support for IPv4/IPv6 ECN marking of SCTP packets (per color). */ + int mark_ip_ecn_sctp_supported[RTE_TM_COLORS]; + + /** Support for IPv4/IPv6 DSCP packet marking (per color). */ + int mark_ip_dscp_supported[RTE_TM_COLORS]; + + /** Set of supported dynamic update operations. + * @see enum rte_tm_dynamic_update_type + */ + uint64_t dynamic_update_mask; + + /** Set of supported statistics counter types. + * @see enum rte_tm_stats_type + */ + uint64_t stats_mask; +}; + +/** + * Traffic manager level capabilities + */ +struct rte_tm_level_capabilities { + /** Maximum number of nodes for the current hierarchy level. */ + uint32_t n_nodes_max; + + /** Maximum number of non-leaf nodes for the current hierarchy level. + * The value of 0 indicates that current level only supports leaf + * nodes. The maximum value is *n_nodes_max*. + */ + uint32_t n_nodes_nonleaf_max; + + /** Maximum number of leaf nodes for the current hierarchy level. The + * value of 0 indicates that current level only supports non-leaf + * nodes. The maximum value is *n_nodes_max*. + */ + uint32_t n_nodes_leaf_max; + + /** When non-zero, this flag indicates that all the non-leaf nodes on + * this level have identical capability set. Valid only when + * *n_nodes_nonleaf_max* is non-zero. + */ + int non_leaf_nodes_identical; + + /** When non-zero, this flag indicates that all the leaf nodes on this + * level have identical capability set. Valid only when + * *n_nodes_leaf_max* is non-zero. + */ + int leaf_nodes_identical; + + RTE_STD_C11 + union { + /** Items valid only for the non-leaf nodes on this level. */ + struct { + /** Private shaper support. When non-zero, it indicates + * there is at least one non-leaf node on this level + * with private shaper support, which may not be the + * case for all the non-leaf nodes on this level. + */ + int shaper_private_supported; + + /** Dual rate support for private shaper. Valid only + * when private shaper is supported for the non-leaf + * nodes on the current level. When non-zero, it + * indicates there is at least one non-leaf node on this + * level with dual rate private shaper support, which + * may not be the case for all the non-leaf nodes on + * this level. + */ + int shaper_private_dual_rate_supported; + + /** Minimum committed/peak rate (bytes per second) for + * private shapers of the non-leaf nodes of this level. + * Valid only when private shaper is supported on this + * level. + */ + uint64_t shaper_private_rate_min; + + /** Maximum committed/peak rate (bytes per second) for + * private shapers of the non-leaf nodes on this level. + * Valid only when private shaper is supported on this + * level. + */ + uint64_t shaper_private_rate_max; + + /** Maximum number of shared shapers that any non-leaf + * node on this level can be part of. The value of zero + * indicates that shared shapers are not supported by + * the non-leaf nodes on this level. When non-zero, it + * indicates there is at least one non-leaf node on this + * level that meets this condition, which may not be the + * case for all the non-leaf nodes on this level. + */ + uint32_t shaper_shared_n_max; + + /** Maximum number of children nodes. This parameter + * indicates that there is at least one non-leaf node on + * this level that can be configured with this many + * children nodes, which might not be true for all the + * non-leaf nodes on this level. + */ + uint32_t sched_n_children_max; + + /** Maximum number of supported priority levels. This + * parameter indicates that there is at least one + * non-leaf node on this level that can be configured + * with this many priority levels for managing its + * children nodes, which might not be true for all the + * non-leaf nodes on this level. The value of zero is + * invalid. The value of 1 indicates that only priority + * 0 is supported, which essentially means that Strict + * Priority (SP) algorithm is not supported on this + * level. + */ + uint32_t sched_sp_n_priorities_max; + + /** Maximum number of sibling nodes that can have the + * same priority at any given time, i.e. maximum size of + * the WFQ sibling node group. This parameter indicates + * there is at least one non-leaf node on this level + * that meets this condition, which may not be true for + * all the non-leaf nodes on this level. The value of + * zero is invalid. The value of 1 indicates that WFQ + * algorithm is not supported on this level. The maximum + * value is *sched_n_children_max*. + */ + uint32_t sched_wfq_n_children_per_group_max; + + /** Maximum number of priority levels that can have + * more than one child node at any given time, i.e. + * maximum number of WFQ sibling node groups that + * have two or more members. This parameter indicates + * there is at least one non-leaf node on this level + * that meets this condition, which might not be true + * for all the non-leaf nodes. The value of zero states + * that WFQ algorithm is not supported on this level. + * The value of 1 indicates that + * (*sched_sp_n_priorities_max* - 1) priority levels on + * this level have at most one child node, so there can + * be only one priority level with two or more sibling + * nodes making up a WFQ group on this level. The + * maximum value is: + * min(floor(*sched_n_children_max* / 2), + * *sched_sp_n_priorities_max*). + */ + uint32_t sched_wfq_n_groups_max; + + /** Maximum WFQ weight. The value of 1 indicates that + * all sibling nodes on this level with same priority + * have the same WFQ weight, so on this level WFQ is + * reduced to FQ. + */ + uint32_t sched_wfq_weight_max; + + /** Mask of statistics counter types supported by the + * non-leaf nodes on this level. Every supported + * statistics counter type is supported by at least one + * non-leaf node on this level, which may not be true + * for all the non-leaf nodes on this level. + * @see enum rte_tm_stats_type + */ + uint64_t stats_mask; + } nonleaf; + + /** Items valid only for the leaf nodes on this level. */ + struct { + /** Private shaper support. When non-zero, it indicates + * there is at least one leaf node on this level with + * private shaper support, which may not be the case for + * all the leaf nodes on this level. + */ + int shaper_private_supported; + + /** Dual rate support for private shaper. Valid only + * when private shaper is supported for the leaf nodes + * on this level. When non-zero, it indicates there is + * at least one leaf node on this level with dual rate + * private shaper support, which may not be the case for + * all the leaf nodes on this level. + */ + int shaper_private_dual_rate_supported; + + /** Minimum committed/peak rate (bytes per second) for + * private shapers of the leaf nodes of this level. + * Valid only when private shaper is supported for the + * leaf nodes on this level. + */ + uint64_t shaper_private_rate_min; + + /** Maximum committed/peak rate (bytes per second) for + * private shapers of the leaf nodes on this level. + * Valid only when private shaper is supported for the + * leaf nodes on this level. + */ + uint64_t shaper_private_rate_max; + + /** Maximum number of shared shapers that any leaf node + * on this level can be part of. The value of zero + * indicates that shared shapers are not supported by + * the leaf nodes on this level. When non-zero, it + * indicates there is at least one leaf node on this + * level that meets this condition, which may not be the + * case for all the leaf nodes on this level. + */ + uint32_t shaper_shared_n_max; + + /** Head drop algorithm support. When non-zero, this + * parameter indicates that there is at least one leaf + * node on this level that supports the head drop + * algorithm, which might not be true for all the leaf + * nodes on this level. + */ + int cman_head_drop_supported; + + /** Private WRED context support. When non-zero, it + * indicates there is at least one node on this level + * with private WRED context support, which may not be + * true for all the leaf nodes on this level. + */ + int cman_wred_context_private_supported; + + /** Maximum number of shared WRED contexts that any + * leaf node on this level can be part of. The value of + * zero indicates that shared WRED contexts are not + * supported by the leaf nodes on this level. When + * non-zero, it indicates there is at least one leaf + * node on this level that meets this condition, which + * may not be the case for all the leaf nodes on this + * level. + */ + uint32_t cman_wred_context_shared_n_max; + + /** Mask of statistics counter types supported by the + * leaf nodes on this level. Every supported statistics + * counter type is supported by at least one leaf node + * on this level, which may not be true for all the leaf + * nodes on this level. + * @see enum rte_tm_stats_type + */ + uint64_t stats_mask; + } leaf; + }; +}; + +/** + * Traffic manager node capabilities + */ +struct rte_tm_node_capabilities { + /** Private shaper support for the current node. */ + int shaper_private_supported; + + /** Dual rate shaping support for private shaper of current node. + * Valid only when private shaper is supported by the current node. + */ + int shaper_private_dual_rate_supported; + + /** Minimum committed/peak rate (bytes per second) for private + * shaper of current node. Valid only when private shaper is supported + * by the current node. + */ + uint64_t shaper_private_rate_min; + + /** Maximum committed/peak rate (bytes per second) for private + * shaper of current node. Valid only when private shaper is supported + * by the current node. + */ + uint64_t shaper_private_rate_max; + + /** Maximum number of shared shapers the current node can be part of. + * The value of zero indicates that shared shapers are not supported by + * the current node. + */ + uint32_t shaper_shared_n_max; + + RTE_STD_C11 + union { + /** Items valid only for non-leaf nodes. */ + struct { + /** Maximum number of children nodes. */ + uint32_t sched_n_children_max; + + /** Maximum number of supported priority levels. The + * value of zero is invalid. The value of 1 indicates + * that only priority 0 is supported, which essentially + * means that Strict Priority (SP) algorithm is not + * supported. + */ + uint32_t sched_sp_n_priorities_max; + + /** Maximum number of sibling nodes that can have the + * same priority at any given time, i.e. maximum size + * of the WFQ sibling node group. The value of zero + * is invalid. The value of 1 indicates that WFQ + * algorithm is not supported. The maximum value is + * *sched_n_children_max*. + */ + uint32_t sched_wfq_n_children_per_group_max; + + /** Maximum number of priority levels that can have + * more than one child node at any given time, i.e. + * maximum number of WFQ sibling node groups that have + * two or more members. The value of zero states that + * WFQ algorithm is not supported. The value of 1 + * indicates that (*sched_sp_n_priorities_max* - 1) + * priority levels have at most one child node, so there + * can be only one priority level with two or more + * sibling nodes making up a WFQ group. The maximum + * value is: min(floor(*sched_n_children_max* / 2), + * *sched_sp_n_priorities_max*). + */ + uint32_t sched_wfq_n_groups_max; + + /** Maximum WFQ weight. The value of 1 indicates that + * all sibling nodes with same priority have the same + * WFQ weight, so WFQ is reduced to FQ. + */ + uint32_t sched_wfq_weight_max; + } nonleaf; + + /** Items valid only for leaf nodes. */ + struct { + /** Head drop algorithm support for current node. */ + int cman_head_drop_supported; + + /** Private WRED context support for current node. */ + int cman_wred_context_private_supported; + + /** Maximum number of shared WRED contexts the current + * node can be part of. The value of zero indicates that + * shared WRED contexts are not supported by the current + * node. + */ + uint32_t cman_wred_context_shared_n_max; + } leaf; + }; + + /** Mask of statistics counter types supported by the current node. + * @see enum rte_tm_stats_type + */ + uint64_t stats_mask; +}; + +/** + * Congestion management (CMAN) mode + * + * This is used for controlling the admission of packets into a packet queue or + * group of packet queues on congestion. On request of writing a new packet + * into the current queue while the queue is full, the *tail drop* algorithm + * drops the new packet while leaving the queue unmodified, as opposed to *head + * drop* algorithm, which drops the packet at the head of the queue (the oldest + * packet waiting in the queue) and admits the new packet at the tail of the + * queue. + * + * The *Random Early Detection (RED)* algorithm works by proactively dropping + * more and more input packets as the queue occupancy builds up. When the queue + * is full or almost full, RED effectively works as *tail drop*. The *Weighted + * RED* algorithm uses a separate set of RED thresholds for each packet color. + */ +enum rte_tm_cman_mode { + RTE_TM_CMAN_TAIL_DROP = 0, /**< Tail drop */ + RTE_TM_CMAN_HEAD_DROP, /**< Head drop */ + RTE_TM_CMAN_WRED, /**< Weighted Random Early Detection (WRED) */ +}; + +/** + * Random Early Detection (RED) profile + */ +struct rte_tm_red_params { + /** Minimum queue threshold */ + uint16_t min_th; + + /** Maximum queue threshold */ + uint16_t max_th; + + /** Inverse of packet marking probability maximum value (maxp), i.e. + * maxp_inv = 1 / maxp + */ + uint16_t maxp_inv; + + /** Negated log2 of queue weight (wq), i.e. wq = 1 / (2 ^ wq_log2) */ + uint16_t wq_log2; +}; + +/** + * Weighted RED (WRED) profile + * + * Multiple WRED contexts can share the same WRED profile. Each leaf node with + * WRED enabled as its congestion management mode has zero or one private WRED + * context (only one leaf node using it) and/or zero, one or several shared + * WRED contexts (multiple leaf nodes use the same WRED context). A private + * WRED context is used to perform congestion management for a single leaf + * node, while a shared WRED context is used to perform congestion management + * for a group of leaf nodes. + */ +struct rte_tm_wred_params { + /** One set of RED parameters per packet color */ + struct rte_tm_red_params red_params[RTE_TM_COLORS]; +}; + +/** + * Token bucket + */ +struct rte_tm_token_bucket { + /** Token bucket rate (bytes per second) */ + uint64_t rate; + + /** Token bucket size (bytes), a.k.a. max burst size */ + uint64_t size; +}; + +/** + * Shaper (rate limiter) profile + * + * Multiple shaper instances can share the same shaper profile. Each node has + * zero or one private shaper (only one node using it) and/or zero, one or + * several shared shapers (multiple nodes use the same shaper instance). + * A private shaper is used to perform traffic shaping for a single node, while + * a shared shaper is used to perform traffic shaping for a group of nodes. + * + * Single rate shapers use a single token bucket. A single rate shaper can be + * configured by setting the rate of the committed bucket to zero, which + * effectively disables this bucket. The peak bucket is used to limit the rate + * and the burst size for the current shaper. + * + * Dual rate shapers use both the committed and the peak token buckets. The + * rate of the peak bucket has to be bigger than zero, as well as greater than + * or equal to the rate of the committed bucket. + */ +struct rte_tm_shaper_params { + /** Committed token bucket */ + struct rte_tm_token_bucket committed; + + /** Peak token bucket */ + struct rte_tm_token_bucket peak; + + /** Signed value to be added to the length of each packet for the + * purpose of shaping. Can be used to correct the packet length with + * the framing overhead bytes that are also consumed on the wire (e.g. + * RTE_TM_ETH_FRAMING_OVERHEAD_FCS). + */ + int32_t pkt_length_adjust; +}; + +/** + * Node parameters + * + * Each non-leaf node has multiple inputs (its children nodes) and single output + * (which is input to its parent node). It arbitrates its inputs using Strict + * Priority (SP) and Weighted Fair Queuing (WFQ) algorithms to schedule input + * packets to its output while observing its shaping (rate limiting) + * constraints. + * + * Algorithms such as Weighted Round Robin (WRR), Byte-level WRR, Deficit WRR + * (DWRR), etc. are considered approximations of the WFQ ideal and are + * assimilated to WFQ, although an associated implementation-dependent trade-off + * on accuracy, performance and resource usage might exist. + * + * Children nodes with different priorities are scheduled using the SP algorithm + * based on their priority, with zero (0) as the highest priority. Children with + * the same priority are scheduled using the WFQ algorithm according to their + * weights. The WFQ weight of a given child node is relative to the sum of the + * weights of all its sibling nodes that have the same priority, with one (1) as + * the lowest weight. For each SP priority, the WFQ weight mode can be set as + * either byte-based or packet-based. + * + * Each leaf node sits on top of a TX queue of the current Ethernet port. Hence, + * the leaf nodes are predefined, with their node IDs set to 0 .. (N-1), where N + * is the number of TX queues configured for the current Ethernet port. The + * non-leaf nodes have their IDs generated by the application. + */ +struct rte_tm_node_params { + /** Shaper profile for the private shaper. The absence of the private + * shaper for the current node is indicated by setting this parameter + * to RTE_TM_SHAPER_PROFILE_ID_NONE. + */ + uint32_t shaper_profile_id; + + /** User allocated array of valid shared shaper IDs. */ + uint32_t *shared_shaper_id; + + /** Number of shared shaper IDs in the *shared_shaper_id* array. */ + uint32_t n_shared_shapers; + + RTE_STD_C11 + union { + /** Parameters only valid for non-leaf nodes. */ + struct { + /** WFQ weight mode for each SP priority. When NULL, it + * indicates that WFQ is to be used for all priorities. + * When non-NULL, it points to a pre-allocated array of + * *n_sp_priorities* values, with non-zero value for + * byte-mode and zero for packet-mode. + */ + int *wfq_weight_mode; + + /** Number of SP priorities. */ + uint32_t n_sp_priorities; + } nonleaf; + + /** Parameters only valid for leaf nodes. */ + struct { + /** Congestion management mode */ + enum rte_tm_cman_mode cman; + + /** WRED parameters (only valid when *cman* is set to + * WRED). + */ + struct { + /** WRED profile for private WRED context. The + * absence of a private WRED context for the + * current leaf node is indicated by value + * RTE_TM_WRED_PROFILE_ID_NONE. + */ + uint32_t wred_profile_id; + + /** User allocated array of shared WRED context + * IDs. When set to NULL, it indicates that the + * current leaf node should not currently be + * part of any shared WRED contexts. + */ + uint32_t *shared_wred_context_id; + + /** Number of elements in the + * *shared_wred_context_id* array. Only valid + * when *shared_wred_context_id* is non-NULL, + * in which case it should be non-zero. + */ + uint32_t n_shared_wred_contexts; + } wred; + } leaf; + }; + + /** Mask of statistics counter types to be enabled for this node. This + * needs to be a subset of the statistics counter types available for + * the current node. Any statistics counter type not included in this + * set is to be disabled for the current node. + * @see enum rte_tm_stats_type + */ + uint64_t stats_mask; +}; + +/** + * Verbose error types. + * + * Most of them provide the type of the object referenced by struct + * rte_tm_error::cause. + */ +enum rte_tm_error_type { + RTE_TM_ERROR_TYPE_NONE, /**< No error. */ + RTE_TM_ERROR_TYPE_UNSPECIFIED, /**< Cause unspecified. */ + RTE_TM_ERROR_TYPE_CAPABILITIES, + RTE_TM_ERROR_TYPE_LEVEL_ID, + RTE_TM_ERROR_TYPE_WRED_PROFILE, + RTE_TM_ERROR_TYPE_WRED_PROFILE_GREEN, + RTE_TM_ERROR_TYPE_WRED_PROFILE_YELLOW, + RTE_TM_ERROR_TYPE_WRED_PROFILE_RED, + RTE_TM_ERROR_TYPE_WRED_PROFILE_ID, + RTE_TM_ERROR_TYPE_SHARED_WRED_CONTEXT_ID, + RTE_TM_ERROR_TYPE_SHAPER_PROFILE, + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_COMMITTED_RATE, + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_COMMITTED_SIZE, + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_PEAK_RATE, + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_PEAK_SIZE, + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_PKT_ADJUST_LEN, + RTE_TM_ERROR_TYPE_SHAPER_PROFILE_ID, + RTE_TM_ERROR_TYPE_SHARED_SHAPER_ID, + RTE_TM_ERROR_TYPE_NODE_PARENT_NODE_ID, + RTE_TM_ERROR_TYPE_NODE_PRIORITY, + RTE_TM_ERROR_TYPE_NODE_WEIGHT, + RTE_TM_ERROR_TYPE_NODE_PARAMS, + RTE_TM_ERROR_TYPE_NODE_PARAMS_SHAPER_PROFILE_ID, + RTE_TM_ERROR_TYPE_NODE_PARAMS_SHARED_SHAPER_ID, + RTE_TM_ERROR_TYPE_NODE_PARAMS_N_SHARED_SHAPERS, + RTE_TM_ERROR_TYPE_NODE_PARAMS_WFQ_WEIGHT_MODE, + RTE_TM_ERROR_TYPE_NODE_PARAMS_N_SP_PRIORITIES, + RTE_TM_ERROR_TYPE_NODE_PARAMS_CMAN, + RTE_TM_ERROR_TYPE_NODE_PARAMS_WRED_PROFILE_ID, + RTE_TM_ERROR_TYPE_NODE_PARAMS_SHARED_WRED_CONTEXT_ID, + RTE_TM_ERROR_TYPE_NODE_PARAMS_N_SHARED_WRED_CONTEXTS, + RTE_TM_ERROR_TYPE_NODE_PARAMS_STATS, + RTE_TM_ERROR_TYPE_NODE_ID, +}; + +/** + * Verbose error structure definition. + * + * This object is normally allocated by applications and set by PMDs, the + * message points to a constant string which does not need to be freed by + * the application, however its pointer can be considered valid only as long + * as its associated DPDK port remains configured. Closing the underlying + * device or unloading the PMD invalidates it. + * + * Both cause and message may be NULL regardless of the error type. + */ +struct rte_tm_error { + enum rte_tm_error_type type; /**< Cause field and error type. */ + const void *cause; /**< Object responsible for the error. */ + const char *message; /**< Human-readable error message. */ +}; + +/** + * Traffic manager get number of leaf nodes + * + * Each leaf node sits on on top of a TX queue of the current Ethernet port. + * Therefore, the set of leaf nodes is predefined, their number is always equal + * to N (where N is the number of TX queues configured for the current port) + * and their IDs are 0 .. (N-1). + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[out] n_leaf_nodes + * Number of leaf nodes for the current port. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + */ +int +rte_tm_get_number_of_leaf_nodes(uint8_t port_id, + uint32_t *n_leaf_nodes, + struct rte_tm_error *error); + +/** + * Traffic manager node ID validate and type (i.e. leaf or non-leaf) get + * + * The leaf nodes have predefined IDs in the range of 0 .. (N-1), where N is + * the number of TX queues of the current Ethernet port. The non-leaf nodes + * have their IDs generated by the application outside of the above range, + * which is reserved for leaf nodes. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID value. Needs to be valid. + * @param[out] is_leaf + * Set to non-zero value when node is leaf and to zero otherwise (non-leaf). + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + */ +int +rte_tm_node_type_get(uint8_t port_id, + uint32_t node_id, + int *is_leaf, + struct rte_tm_error *error); + +/** + * Traffic manager capabilities get + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[out] cap + * Traffic manager capabilities. Needs to be pre-allocated and valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + */ +int +rte_tm_capabilities_get(uint8_t port_id, + struct rte_tm_capabilities *cap, + struct rte_tm_error *error); + +/** + * Traffic manager level capabilities get + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] level_id + * The hierarchy level identifier. The value of 0 identifies the level of the + * root node. + * @param[out] cap + * Traffic manager level capabilities. Needs to be pre-allocated and valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + */ +int +rte_tm_level_capabilities_get(uint8_t port_id, + uint32_t level_id, + struct rte_tm_level_capabilities *cap, + struct rte_tm_error *error); + +/** + * Traffic manager node capabilities get + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[out] cap + * Traffic manager node capabilities. Needs to be pre-allocated and valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + */ +int +rte_tm_node_capabilities_get(uint8_t port_id, + uint32_t node_id, + struct rte_tm_node_capabilities *cap, + struct rte_tm_error *error); + +/** + * Traffic manager WRED profile add + * + * Create a new WRED profile with ID set to *wred_profile_id*. The new profile + * is used to create one or several WRED contexts. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] wred_profile_id + * WRED profile ID for the new profile. Needs to be unused. + * @param[in] profile + * WRED profile parameters. Needs to be pre-allocated and valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::cman_wred_context_n_max + */ +int +rte_tm_wred_profile_add(uint8_t port_id, + uint32_t wred_profile_id, + struct rte_tm_wred_params *profile, + struct rte_tm_error *error); + +/** + * Traffic manager WRED profile delete + * + * Delete an existing WRED profile. This operation fails when there is + * currently at least one user (i.e. WRED context) of this WRED profile. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] wred_profile_id + * WRED profile ID. Needs to be the valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::cman_wred_context_n_max + */ +int +rte_tm_wred_profile_delete(uint8_t port_id, + uint32_t wred_profile_id, + struct rte_tm_error *error); + +/** + * Traffic manager shared WRED context add or update + * + * When *shared_wred_context_id* is invalid, a new WRED context with this ID is + * created by using the WRED profile identified by *wred_profile_id*. + * + * When *shared_wred_context_id* is valid, this WRED context is no longer using + * the profile previously assigned to it and is updated to use the profile + * identified by *wred_profile_id*. + * + * A valid shared WRED context can be assigned to several hierarchy leaf nodes + * configured to use WRED as the congestion management mode. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] shared_wred_context_id + * Shared WRED context ID + * @param[in] wred_profile_id + * WRED profile ID. Needs to be the valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::cman_wred_context_shared_n_max + */ +int +rte_tm_shared_wred_context_add_update(uint8_t port_id, + uint32_t shared_wred_context_id, + uint32_t wred_profile_id, + struct rte_tm_error *error); + +/** + * Traffic manager shared WRED context delete + * + * Delete an existing shared WRED context. This operation fails when there is + * currently at least one user (i.e. hierarchy leaf node) of this shared WRED + * context. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] shared_wred_context_id + * Shared WRED context ID. Needs to be the valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::cman_wred_context_shared_n_max + */ +int +rte_tm_shared_wred_context_delete(uint8_t port_id, + uint32_t shared_wred_context_id, + struct rte_tm_error *error); + +/** + * Traffic manager shaper profile add + * + * Create a new shaper profile with ID set to *shaper_profile_id*. The new + * shaper profile is used to create one or several shapers. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] shaper_profile_id + * Shaper profile ID for the new profile. Needs to be unused. + * @param[in] profile + * Shaper profile parameters. Needs to be pre-allocated and valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::shaper_n_max + */ +int +rte_tm_shaper_profile_add(uint8_t port_id, + uint32_t shaper_profile_id, + struct rte_tm_shaper_params *profile, + struct rte_tm_error *error); + +/** + * Traffic manager shaper profile delete + * + * Delete an existing shaper profile. This operation fails when there is + * currently at least one user (i.e. shaper) of this shaper profile. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] shaper_profile_id + * Shaper profile ID. Needs to be the valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::shaper_n_max + */ +int +rte_tm_shaper_profile_delete(uint8_t port_id, + uint32_t shaper_profile_id, + struct rte_tm_error *error); + +/** + * Traffic manager shared shaper add or update + * + * When *shared_shaper_id* is not a valid shared shaper ID, a new shared shaper + * with this ID is created using the shaper profile identified by + * *shaper_profile_id*. + * + * When *shared_shaper_id* is a valid shared shaper ID, this shared shaper is + * no longer using the shaper profile previously assigned to it and is updated + * to use the shaper profile identified by *shaper_profile_id*. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] shared_shaper_id + * Shared shaper ID + * @param[in] shaper_profile_id + * Shaper profile ID. Needs to be the valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::shaper_shared_n_max + */ +int +rte_tm_shared_shaper_add_update(uint8_t port_id, + uint32_t shared_shaper_id, + uint32_t shaper_profile_id, + struct rte_tm_error *error); + +/** + * Traffic manager shared shaper delete + * + * Delete an existing shared shaper. This operation fails when there is + * currently at least one user (i.e. hierarchy node) of this shared shaper. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] shared_shaper_id + * Shared shaper ID. Needs to be the valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::shaper_shared_n_max + */ +int +rte_tm_shared_shaper_delete(uint8_t port_id, + uint32_t shared_shaper_id, + struct rte_tm_error *error); + +/** + * Traffic manager node add + * + * Create new node and connect it as child of an existing node. The new node is + * further identified by *node_id*, which needs to be unused by any of the + * existing nodes. The parent node is identified by *parent_node_id*, which + * needs to be the valid ID of an existing non-leaf node. The parent node is + * going to use the provided SP *priority* and WFQ *weight* to schedule its new + * child node. + * + * This function has to be called for both leaf and non-leaf nodes. In the case + * of leaf nodes (i.e. *node_id* is within the range of 0 .. (N-1), with N as + * the number of configured TX queues of the current port), the leaf node is + * configured rather than created (as the set of leaf nodes is predefined) and + * it is also connected as child of an existing node. + * + * The first node that is added becomes the root node and all the nodes that + * are subsequently added have to be added as descendants of the root node. The + * parent of the root node has to be specified as RTE_TM_NODE_ID_NULL and there + * can only be one node with this parent ID (i.e. the root node). Further + * restrictions for root node: needs to be non-leaf, its private shaper profile + * needs to be valid and single rate, cannot use any shared shapers. + * + * When called before rte_tm_hierarchy_commit() invocation, this function is + * typically used to define the initial start-up hierarchy for the port. + * Provided that dynamic hierarchy updates are supported by the current port (as + * advertised in the port capability set), this function can be also called + * after the rte_tm_hierarchy_commit() invocation. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be unused by any of the existing nodes. + * @param[in] parent_node_id + * Parent node ID. Needs to be the valid. + * @param[in] priority + * Node priority. The highest node priority is zero. Used by the SP algorithm + * running on the parent of the current node for scheduling this child node. + * @param[in] weight + * Node weight. The node weight is relative to the weight sum of all siblings + * that have the same priority. The lowest weight is one. Used by the WFQ + * algorithm running on the parent of the current node for scheduling this + * child node. + * @param[in] level_id + * Level ID that should be met by this node. The hierarchy level of the + * current node is already fully specified through its parent node (i.e. the + * level of this node is equal to the level of its parent node plus one), + * therefore the reason for providing this parameter is to enable the + * application to perform step-by-step checking of the node level during + * successive invocations of this function. When not desired, this check can + * be disabled by assigning value RTE_TM_NODE_LEVEL_ID_ANY to this parameter. + * @param[in] params + * Node parameters. Needs to be pre-allocated and valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see rte_tm_hierarchy_commit() + * @see RTE_TM_UPDATE_NODE_ADD_DELETE + * @see RTE_TM_NODE_LEVEL_ID_ANY + * @see struct rte_tm_capabilities + */ +int +rte_tm_node_add(uint8_t port_id, + uint32_t node_id, + uint32_t parent_node_id, + uint32_t priority, + uint32_t weight, + uint32_t level_id, + struct rte_tm_node_params *params, + struct rte_tm_error *error); + +/** + * Traffic manager node delete + * + * Delete an existing node. This operation fails when this node currently has + * at least one user (i.e. child node). + * + * When called before rte_tm_hierarchy_commit() invocation, this function is + * typically used to define the initial start-up hierarchy for the port. + * Provided that dynamic hierarchy updates are supported by the current port (as + * advertised in the port capability set), this function can be also called + * after the rte_tm_hierarchy_commit() invocation. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see RTE_TM_UPDATE_NODE_ADD_DELETE + */ +int +rte_tm_node_delete(uint8_t port_id, + uint32_t node_id, + struct rte_tm_error *error); + +/** + * Traffic manager node suspend + * + * Suspend an existing node. While the node is in suspended state, no packet is + * scheduled from this node and its descendants. The node exits the suspended + * state through the node resume operation. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see rte_tm_node_resume() + * @see RTE_TM_UPDATE_NODE_SUSPEND_RESUME + */ +int +rte_tm_node_suspend(uint8_t port_id, + uint32_t node_id, + struct rte_tm_error *error); + +/** + * Traffic manager node resume + * + * Resume an existing node that is currently in suspended state. The node + * entered the suspended state as result of a previous node suspend operation. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see rte_tm_node_suspend() + * @see RTE_TM_UPDATE_NODE_SUSPEND_RESUME + */ +int +rte_tm_node_resume(uint8_t port_id, + uint32_t node_id, + struct rte_tm_error *error); + +/** + * Traffic manager hierarchy commit + * + * This function is called during the port initialization phase (before the + * Ethernet port is started) to freeze the start-up hierarchy. + * + * This function typically performs the following steps: + * a) It validates the start-up hierarchy that was previously defined for the + * current port through successive rte_tm_node_add() invocations; + * b) Assuming successful validation, it performs all the necessary port + * specific configuration operations to install the specified hierarchy on + * the current port, with immediate effect once the port is started. + * + * This function fails when the currently configured hierarchy is not supported + * by the Ethernet port, in which case the user can abort or try out another + * hierarchy configuration (e.g. a hierarchy with less leaf nodes), which can be + * build from scratch (when *clear_on_fail* is enabled) or by modifying the + * existing hierarchy configuration (when *clear_on_fail* is disabled). + * + * Note that this function can still fail due to other causes (e.g. not enough + * memory available in the system, etc), even though the specified hierarchy is + * supported in principle by the current port. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] clear_on_fail + * On function call failure, hierarchy is cleared when this parameter is + * non-zero and preserved when this parameter is equal to zero. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see rte_tm_node_add() + * @see rte_tm_node_delete() + */ +int +rte_tm_hierarchy_commit(uint8_t port_id, + int clear_on_fail, + struct rte_tm_error *error); + +/** + * Traffic manager node parent update + * + * Restriction for root node: its parent cannot be changed. + * + * This function can only be called after the rte_tm_hierarchy_commit() + * invocation. Its success depends on the port support for this operation, as + * advertised through the port capability set. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[in] parent_node_id + * Node ID for the new parent. Needs to be valid. + * @param[in] priority + * Node priority. The highest node priority is zero. Used by the SP algorithm + * running on the parent of the current node for scheduling this child node. + * @param[in] weight + * Node weight. The node weight is relative to the weight sum of all siblings + * that have the same priority. The lowest weight is zero. Used by the WFQ + * algorithm running on the parent of the current node for scheduling this + * child node. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see RTE_TM_UPDATE_NODE_PARENT_KEEP_LEVEL + * @see RTE_TM_UPDATE_NODE_PARENT_CHANGE_LEVEL + */ +int +rte_tm_node_parent_update(uint8_t port_id, + uint32_t node_id, + uint32_t parent_node_id, + uint32_t priority, + uint32_t weight, + struct rte_tm_error *error); + +/** + * Traffic manager node private shaper update + * + * Restriction for the root node: its private shaper profile needs to be valid + * and single rate. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[in] shaper_profile_id + * Shaper profile ID for the private shaper of the current node. Needs to be + * either valid shaper profile ID or RTE_TM_SHAPER_PROFILE_ID_NONE, with + * the latter disabling the private shaper of the current node. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::shaper_private_n_max + */ +int +rte_tm_node_shaper_update(uint8_t port_id, + uint32_t node_id, + uint32_t shaper_profile_id, + struct rte_tm_error *error); + +/** + * Traffic manager node shared shapers update + * + * Restriction for root node: cannot use any shared rate shapers. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[in] shared_shaper_id + * Shared shaper ID. Needs to be valid. + * @param[in] add + * Set to non-zero value to add this shared shaper to current node or to zero + * to delete this shared shaper from current node. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::shaper_shared_n_max + */ +int +rte_tm_node_shared_shaper_update(uint8_t port_id, + uint32_t node_id, + uint32_t shared_shaper_id, + int add, + struct rte_tm_error *error); + +/** + * Traffic manager node enabled statistics counters update + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[in] stats_mask + * Mask of statistics counter types to be enabled for the current node. This + * needs to be a subset of the statistics counter types available for the + * current node. Any statistics counter type not included in this set is to + * be disabled for the current node. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see enum rte_tm_stats_type + * @see RTE_TM_UPDATE_NODE_STATS + */ +int +rte_tm_node_stats_update(uint8_t port_id, + uint32_t node_id, + uint64_t stats_mask, + struct rte_tm_error *error); + +/** + * Traffic manager node WFQ weight mode update + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid leaf node ID. + * @param[in] wfq_weight_mode + * WFQ weight mode for each SP priority. When NULL, it indicates that WFQ is + * to be used for all priorities. When non-NULL, it points to a pre-allocated + * array of *n_sp_priorities* values, with non-zero value for byte-mode and + * zero for packet-mode. + * @param[in] n_sp_priorities + * Number of SP priorities. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see RTE_TM_UPDATE_NODE_WFQ_WEIGHT_MODE + * @see RTE_TM_UPDATE_NODE_N_SP_PRIORITIES + */ +int +rte_tm_node_wfq_weight_mode_update(uint8_t port_id, + uint32_t node_id, + int *wfq_weight_mode, + uint32_t n_sp_priorities, + struct rte_tm_error *error); + +/** + * Traffic manager node congestion management mode update + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid leaf node ID. + * @param[in] cman + * Congestion management mode. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see RTE_TM_UPDATE_NODE_CMAN + */ +int +rte_tm_node_cman_update(uint8_t port_id, + uint32_t node_id, + enum rte_tm_cman_mode cman, + struct rte_tm_error *error); + +/** + * Traffic manager node private WRED context update + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid leaf node ID. + * @param[in] wred_profile_id + * WRED profile ID for the private WRED context of the current node. Needs to + * be either valid WRED profile ID or RTE_TM_WRED_PROFILE_ID_NONE, with the + * latter disabling the private WRED context of the current node. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::cman_wred_context_private_n_max +*/ +int +rte_tm_node_wred_context_update(uint8_t port_id, + uint32_t node_id, + uint32_t wred_profile_id, + struct rte_tm_error *error); + +/** + * Traffic manager node shared WRED context update + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid leaf node ID. + * @param[in] shared_wred_context_id + * Shared WRED context ID. Needs to be valid. + * @param[in] add + * Set to non-zero value to add this shared WRED context to current node or + * to zero to delete this shared WRED context from current node. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::cman_wred_context_shared_n_max + */ +int +rte_tm_node_shared_wred_context_update(uint8_t port_id, + uint32_t node_id, + uint32_t shared_wred_context_id, + int add, + struct rte_tm_error *error); + +/** + * Traffic manager node statistics counters read + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] node_id + * Node ID. Needs to be valid. + * @param[out] stats + * When non-NULL, it contains the current value for the statistics counters + * enabled for the current node. + * @param[out] stats_mask + * When non-NULL, it contains the mask of statistics counter types that are + * currently enabled for this node, indicating which of the counters + * retrieved with the *stats* structure are valid. + * @param[in] clear + * When this parameter has a non-zero value, the statistics counters are + * cleared (i.e. set to zero) immediately after they have been read, + * otherwise the statistics counters are left untouched. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see enum rte_tm_stats_type + */ +int +rte_tm_node_stats_read(uint8_t port_id, + uint32_t node_id, + struct rte_tm_node_stats *stats, + uint64_t *stats_mask, + int clear, + struct rte_tm_error *error); + +/** + * Traffic manager packet marking - VLAN DEI (IEEE 802.1Q) + * + * IEEE 802.1p maps the traffic class to the VLAN Priority Code Point (PCP) + * field (3 bits), while IEEE 802.1q maps the drop priority to the VLAN Drop + * Eligible Indicator (DEI) field (1 bit), which was previously named Canonical + * Format Indicator (CFI). + * + * All VLAN frames of a given color get their DEI bit set if marking is enabled + * for this color; otherwise, their DEI bit is left as is (either set or not). + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] mark_green + * Set to non-zero value to enable marking of green packets and to zero to + * disable it. + * @param[in] mark_yellow + * Set to non-zero value to enable marking of yellow packets and to zero to + * disable it. + * @param[in] mark_red + * Set to non-zero value to enable marking of red packets and to zero to + * disable it. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::mark_vlan_dei_supported + */ +int +rte_tm_mark_vlan_dei(uint8_t port_id, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error); + +/** + * Traffic manager packet marking - IPv4 / IPv6 ECN (IETF RFC 3168) + * + * IETF RFCs 2474 and 3168 reorganize the IPv4 Type of Service (TOS) field + * (8 bits) and the IPv6 Traffic Class (TC) field (8 bits) into Differentiated + * Services Codepoint (DSCP) field (6 bits) and Explicit Congestion + * Notification (ECN) field (2 bits). The DSCP field is typically used to + * encode the traffic class and/or drop priority (RFC 2597), while the ECN + * field is used by RFC 3168 to implement a congestion notification mechanism + * to be leveraged by transport layer protocols such as TCP and SCTP that have + * congestion control mechanisms. + * + * When congestion is experienced, as alternative to dropping the packet, + * routers can change the ECN field of input packets from 2'b01 or 2'b10 + * (values indicating that source endpoint is ECN-capable) to 2'b11 (meaning + * that congestion is experienced). The destination endpoint can use the + * ECN-Echo (ECE) TCP flag to relay the congestion indication back to the + * source endpoint, which acknowledges it back to the destination endpoint with + * the Congestion Window Reduced (CWR) TCP flag. + * + * All IPv4/IPv6 packets of a given color with ECN set to 2’b01 or 2’b10 + * carrying TCP or SCTP have their ECN set to 2’b11 if the marking feature is + * enabled for the current color, otherwise the ECN field is left as is. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] mark_green + * Set to non-zero value to enable marking of green packets and to zero to + * disable it. + * @param[in] mark_yellow + * Set to non-zero value to enable marking of yellow packets and to zero to + * disable it. + * @param[in] mark_red + * Set to non-zero value to enable marking of red packets and to zero to + * disable it. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::mark_ip_ecn_tcp_supported + * @see struct rte_tm_capabilities::mark_ip_ecn_sctp_supported + */ +int +rte_tm_mark_ip_ecn(uint8_t port_id, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error); + +/** + * Traffic manager packet marking - IPv4 / IPv6 DSCP (IETF RFC 2597) + * + * IETF RFC 2597 maps the traffic class and the drop priority to the IPv4/IPv6 + * Differentiated Services Codepoint (DSCP) field (6 bits). Here are the DSCP + * values proposed by this RFC: + * + * 
                   Class 1    Class 2    Class 3    Class 4
+ * 
                 +----------+----------+----------+----------+
+ * 
Low Drop Prec    |  001010  |  010010  |  011010  |  100010  |
+ * 
Medium Drop Prec |  001100  |  010100  |  011100  |  100100  |
+ * 
High Drop Prec   |  001110  |  010110  |  011110  |  100110  |
+ * 
                 +----------+----------+----------+----------+
+ * + * There are 4 traffic classes (classes 1 .. 4) encoded by DSCP bits 1 and 2, + * as well as 3 drop priorities (low/medium/high) encoded by DSCP bits 3 and 4. + * + * All IPv4/IPv6 packets have their color marked into DSCP bits 3 and 4 as + * follows: green mapped to Low Drop Precedence (2’b01), yellow to Medium + * (2’b10) and red to High (2’b11). Marking needs to be explicitly enabled + * for each color; when not enabled for a given color, the DSCP field of all + * packets with that color is left as is. + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[in] mark_green + * Set to non-zero value to enable marking of green packets and to zero to + * disable it. + * @param[in] mark_yellow + * Set to non-zero value to enable marking of yellow packets and to zero to + * disable it. + * @param[in] mark_red + * Set to non-zero value to enable marking of red packets and to zero to + * disable it. + * @param[out] error + * Error details. Filled in only on error, when not NULL. + * @return + * 0 on success, non-zero error code otherwise. + * + * @see struct rte_tm_capabilities::mark_ip_dscp_supported + */ +int +rte_tm_mark_ip_dscp(uint8_t port_id, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error); + +#ifdef __cplusplus +} +#endif + +#endif /* __INCLUDE_RTE_TM_H__ */ diff --git a/lib/librte_ether/rte_tm_driver.h b/lib/librte_ether/rte_tm_driver.h new file mode 100644 index 00000000..a5b698fe --- /dev/null +++ b/lib/librte_ether/rte_tm_driver.h @@ -0,0 +1,366 @@ +/*- + * BSD LICENSE + * + * Copyright(c) 2017 Intel Corporation. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * * Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * * Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * * Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#ifndef __INCLUDE_RTE_TM_DRIVER_H__ +#define __INCLUDE_RTE_TM_DRIVER_H__ + +/** + * @file + * RTE Generic Traffic Manager API (Driver Side) + * + * This file provides implementation helpers for internal use by PMDs, they + * are not intended to be exposed to applications and are not subject to ABI + * versioning. + */ + +#include + +#include +#include "rte_ethdev.h" +#include "rte_tm.h" + +#ifdef __cplusplus +extern "C" { +#endif + +/** @internal Traffic manager node ID validate and type get */ +typedef int (*rte_tm_node_type_get_t)(struct rte_eth_dev *dev, + uint32_t node_id, + int *is_leaf, + struct rte_tm_error *error); + +/** @internal Traffic manager capabilities get */ +typedef int (*rte_tm_capabilities_get_t)(struct rte_eth_dev *dev, + struct rte_tm_capabilities *cap, + struct rte_tm_error *error); + +/** @internal Traffic manager level capabilities get */ +typedef int (*rte_tm_level_capabilities_get_t)(struct rte_eth_dev *dev, + uint32_t level_id, + struct rte_tm_level_capabilities *cap, + struct rte_tm_error *error); + +/** @internal Traffic manager node capabilities get */ +typedef int (*rte_tm_node_capabilities_get_t)(struct rte_eth_dev *dev, + uint32_t node_id, + struct rte_tm_node_capabilities *cap, + struct rte_tm_error *error); + +/** @internal Traffic manager WRED profile add */ +typedef int (*rte_tm_wred_profile_add_t)(struct rte_eth_dev *dev, + uint32_t wred_profile_id, + struct rte_tm_wred_params *profile, + struct rte_tm_error *error); + +/** @internal Traffic manager WRED profile delete */ +typedef int (*rte_tm_wred_profile_delete_t)(struct rte_eth_dev *dev, + uint32_t wred_profile_id, + struct rte_tm_error *error); + +/** @internal Traffic manager shared WRED context add */ +typedef int (*rte_tm_shared_wred_context_add_update_t)( + struct rte_eth_dev *dev, + uint32_t shared_wred_context_id, + uint32_t wred_profile_id, + struct rte_tm_error *error); + +/** @internal Traffic manager shared WRED context delete */ +typedef int (*rte_tm_shared_wred_context_delete_t)( + struct rte_eth_dev *dev, + uint32_t shared_wred_context_id, + struct rte_tm_error *error); + +/** @internal Traffic manager shaper profile add */ +typedef int (*rte_tm_shaper_profile_add_t)(struct rte_eth_dev *dev, + uint32_t shaper_profile_id, + struct rte_tm_shaper_params *profile, + struct rte_tm_error *error); + +/** @internal Traffic manager shaper profile delete */ +typedef int (*rte_tm_shaper_profile_delete_t)(struct rte_eth_dev *dev, + uint32_t shaper_profile_id, + struct rte_tm_error *error); + +/** @internal Traffic manager shared shaper add/update */ +typedef int (*rte_tm_shared_shaper_add_update_t)(struct rte_eth_dev *dev, + uint32_t shared_shaper_id, + uint32_t shaper_profile_id, + struct rte_tm_error *error); + +/** @internal Traffic manager shared shaper delete */ +typedef int (*rte_tm_shared_shaper_delete_t)(struct rte_eth_dev *dev, + uint32_t shared_shaper_id, + struct rte_tm_error *error); + +/** @internal Traffic manager node add */ +typedef int (*rte_tm_node_add_t)(struct rte_eth_dev *dev, + uint32_t node_id, + uint32_t parent_node_id, + uint32_t priority, + uint32_t weight, + uint32_t level_id, + struct rte_tm_node_params *params, + struct rte_tm_error *error); + +/** @internal Traffic manager node delete */ +typedef int (*rte_tm_node_delete_t)(struct rte_eth_dev *dev, + uint32_t node_id, + struct rte_tm_error *error); + +/** @internal Traffic manager node suspend */ +typedef int (*rte_tm_node_suspend_t)(struct rte_eth_dev *dev, + uint32_t node_id, + struct rte_tm_error *error); + +/** @internal Traffic manager node resume */ +typedef int (*rte_tm_node_resume_t)(struct rte_eth_dev *dev, + uint32_t node_id, + struct rte_tm_error *error); + +/** @internal Traffic manager hierarchy commit */ +typedef int (*rte_tm_hierarchy_commit_t)(struct rte_eth_dev *dev, + int clear_on_fail, + struct rte_tm_error *error); + +/** @internal Traffic manager node parent update */ +typedef int (*rte_tm_node_parent_update_t)(struct rte_eth_dev *dev, + uint32_t node_id, + uint32_t parent_node_id, + uint32_t priority, + uint32_t weight, + struct rte_tm_error *error); + +/** @internal Traffic manager node shaper update */ +typedef int (*rte_tm_node_shaper_update_t)(struct rte_eth_dev *dev, + uint32_t node_id, + uint32_t shaper_profile_id, + struct rte_tm_error *error); + +/** @internal Traffic manager node shaper update */ +typedef int (*rte_tm_node_shared_shaper_update_t)(struct rte_eth_dev *dev, + uint32_t node_id, + uint32_t shared_shaper_id, + int32_t add, + struct rte_tm_error *error); + +/** @internal Traffic manager node stats update */ +typedef int (*rte_tm_node_stats_update_t)(struct rte_eth_dev *dev, + uint32_t node_id, + uint64_t stats_mask, + struct rte_tm_error *error); + +/** @internal Traffic manager node WFQ weight mode update */ +typedef int (*rte_tm_node_wfq_weight_mode_update_t)( + struct rte_eth_dev *dev, + uint32_t node_id, + int *wfq_weigth_mode, + uint32_t n_sp_priorities, + struct rte_tm_error *error); + +/** @internal Traffic manager node congestion management mode update */ +typedef int (*rte_tm_node_cman_update_t)(struct rte_eth_dev *dev, + uint32_t node_id, + enum rte_tm_cman_mode cman, + struct rte_tm_error *error); + +/** @internal Traffic manager node WRED context update */ +typedef int (*rte_tm_node_wred_context_update_t)( + struct rte_eth_dev *dev, + uint32_t node_id, + uint32_t wred_profile_id, + struct rte_tm_error *error); + +/** @internal Traffic manager node WRED context update */ +typedef int (*rte_tm_node_shared_wred_context_update_t)( + struct rte_eth_dev *dev, + uint32_t node_id, + uint32_t shared_wred_context_id, + int add, + struct rte_tm_error *error); + +/** @internal Traffic manager read stats counters for specific node */ +typedef int (*rte_tm_node_stats_read_t)(struct rte_eth_dev *dev, + uint32_t node_id, + struct rte_tm_node_stats *stats, + uint64_t *stats_mask, + int clear, + struct rte_tm_error *error); + +/** @internal Traffic manager packet marking - VLAN DEI */ +typedef int (*rte_tm_mark_vlan_dei_t)(struct rte_eth_dev *dev, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error); + +/** @internal Traffic manager packet marking - IPv4/IPv6 ECN */ +typedef int (*rte_tm_mark_ip_ecn_t)(struct rte_eth_dev *dev, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error); + +/** @internal Traffic manager packet marking - IPv4/IPv6 DSCP */ +typedef int (*rte_tm_mark_ip_dscp_t)(struct rte_eth_dev *dev, + int mark_green, + int mark_yellow, + int mark_red, + struct rte_tm_error *error); + +struct rte_tm_ops { + /** Traffic manager node type get */ + rte_tm_node_type_get_t node_type_get; + + /** Traffic manager capabilities_get */ + rte_tm_capabilities_get_t capabilities_get; + /** Traffic manager level capabilities_get */ + rte_tm_level_capabilities_get_t level_capabilities_get; + /** Traffic manager node capabilities get */ + rte_tm_node_capabilities_get_t node_capabilities_get; + + /** Traffic manager WRED profile add */ + rte_tm_wred_profile_add_t wred_profile_add; + /** Traffic manager WRED profile delete */ + rte_tm_wred_profile_delete_t wred_profile_delete; + /** Traffic manager shared WRED context add/update */ + rte_tm_shared_wred_context_add_update_t + shared_wred_context_add_update; + /** Traffic manager shared WRED context delete */ + rte_tm_shared_wred_context_delete_t + shared_wred_context_delete; + + /** Traffic manager shaper profile add */ + rte_tm_shaper_profile_add_t shaper_profile_add; + /** Traffic manager shaper profile delete */ + rte_tm_shaper_profile_delete_t shaper_profile_delete; + /** Traffic manager shared shaper add/update */ + rte_tm_shared_shaper_add_update_t shared_shaper_add_update; + /** Traffic manager shared shaper delete */ + rte_tm_shared_shaper_delete_t shared_shaper_delete; + + /** Traffic manager node add */ + rte_tm_node_add_t node_add; + /** Traffic manager node delete */ + rte_tm_node_delete_t node_delete; + /** Traffic manager node suspend */ + rte_tm_node_suspend_t node_suspend; + /** Traffic manager node resume */ + rte_tm_node_resume_t node_resume; + /** Traffic manager hierarchy commit */ + rte_tm_hierarchy_commit_t hierarchy_commit; + + /** Traffic manager node parent update */ + rte_tm_node_parent_update_t node_parent_update; + /** Traffic manager node shaper update */ + rte_tm_node_shaper_update_t node_shaper_update; + /** Traffic manager node shared shaper update */ + rte_tm_node_shared_shaper_update_t node_shared_shaper_update; + /** Traffic manager node stats update */ + rte_tm_node_stats_update_t node_stats_update; + /** Traffic manager node WFQ weight mode update */ + rte_tm_node_wfq_weight_mode_update_t node_wfq_weight_mode_update; + /** Traffic manager node congestion management mode update */ + rte_tm_node_cman_update_t node_cman_update; + /** Traffic manager node WRED context update */ + rte_tm_node_wred_context_update_t node_wred_context_update; + /** Traffic manager node shared WRED context update */ + rte_tm_node_shared_wred_context_update_t + node_shared_wred_context_update; + /** Traffic manager read statistics counters for current node */ + rte_tm_node_stats_read_t node_stats_read; + + /** Traffic manager packet marking - VLAN DEI */ + rte_tm_mark_vlan_dei_t mark_vlan_dei; + /** Traffic manager packet marking - IPv4/IPv6 ECN */ + rte_tm_mark_ip_ecn_t mark_ip_ecn; + /** Traffic manager packet marking - IPv4/IPv6 DSCP */ + rte_tm_mark_ip_dscp_t mark_ip_dscp; +}; + +/** + * Initialize generic error structure. + * + * This function also sets rte_errno to a given value. + * + * @param[out] error + * Pointer to error structure (may be NULL). + * @param[in] code + * Related error code (rte_errno). + * @param[in] type + * Cause field and error type. + * @param[in] cause + * Object responsible for the error. + * @param[in] message + * Human-readable error message. + * + * @return + * Error code. + */ +static inline int +rte_tm_error_set(struct rte_tm_error *error, + int code, + enum rte_tm_error_type type, + const void *cause, + const char *message) +{ + if (error) { + *error = (struct rte_tm_error){ + .type = type, + .cause = cause, + .message = message, + }; + } + rte_errno = code; + return code; +} + +/** + * Get generic traffic manager operations structure from a port + * + * @param[in] port_id + * The port identifier of the Ethernet device. + * @param[out] error + * Error details + * + * @return + * The traffic manager operations structure associated with port_id on + * success, NULL otherwise. + */ +const struct rte_tm_ops * +rte_tm_ops_get(uint8_t port_id, struct rte_tm_error *error); + +#ifdef __cplusplus +} +#endif + +#endif /* __INCLUDE_RTE_TM_DRIVER_H__ */ -- cgit 1.2.3-korg